Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Temporal and spatial expression patterns of the hsp16 and ubq-1 genes in transgenic C. elegans Durovic, Eve Gabrielle Stingham 1992

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1992_spring_durovic_eve_gabriell.pdf [ 3.89MB ]
Metadata
JSON: 831-1.0086681.json
JSON-LD: 831-1.0086681-ld.json
RDF/XML (Pretty): 831-1.0086681-rdf.xml
RDF/JSON: 831-1.0086681-rdf.json
Turtle: 831-1.0086681-turtle.txt
N-Triples: 831-1.0086681-rdf-ntriples.txt
Original Record: 831-1.0086681-source.json
Full Text
831-1.0086681-fulltext.txt
Citation
831-1.0086681.ris

Full Text

Temporal and spatial expression patterns of the hspl6 and ubq-1 genes intransgenic C. elegans.byEve Gabrielle Stringham DurovicB.Sc., The University of Manitoba, 1984M.Sc., The University of Manitoba, 1987A THESIS SUBMfl’TED IN PARTIAL FULHLLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIESDepartment of Biochemisty-Genetics ProgramWe accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIANovember, 1991Eve G. Stringham Durovic, 1991In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives, It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)__________________________Department of_____________________The University of British ColumbiaVancouver, CanadaDate 44’l/_ 7/9DE-6 (2/88)In presenting this thesis in partial fulfilment of the requirements foran advanced degree at The University of British Columbia, I agreethat the Library shall make it freely available for reference andstudy. I further agree that permission for extensive copying of thisthesis for scholarly purposes may be granted by the Head of myDepartment or by his or her representatives. It is understood thatcopying or publication of this thesis for financial gain shall not beallowed without my written permission.Biochemisty-Genetics ProgramThe University of British Columbia2075 Wesbrook PlaceVancouver, CanadaV6T 1W5Date: 22/11/91AbstractABSTRACTThe expression of the small 16 kDa heat shock protein gene (hspl6 ) family and of thepolyubiquitin encoding gene (ubq-1) in Caenorhabditis elegans has been examined byintroducing IacZ fusions into the nematode by transformation.Transcription of the hspl6-IacZ transgenes was totally heat shock dependent andresulted in the rapid synthesis of detectable levels of -gaIactosidase in most somatic tissues.Although the two hspl6 gene pairs of C. elegans are highly similar within both their coding andnon-coding sequences, quantitative and qualitative differences in the spatial pattern ofexpression between gene pairs were observed. The hspl6-48 promoter was shown to directgreater expression of 3-galactosidase in muscle and hypodermis while the hspl6-41 promoterwas more efficient in intestine and pharyngeal tissue. Transgenes which eliminated onepromoter from a gene pair were expressed- at reduced levels, particularly in post-embryonicstages, suggesting that the heat shock elements (HSEs) in the intergenic region of an hspl6 genepair may act cooperatively to achieve high levels of expression of both genes. The hspl6 genepairs are never constitutively expressed. In addition, their heat inducibility is developmentallyrestricted; they are not heat inducible during gametogenesis or early embryogenesis. Thehspl6 genes represent the first fully inducible system in C. elegans to be characterized in detailat the molecular level, and the promoters of these genes should be useful for studying the actionof tissue or developmentally regulated genes in this organism.Animals carrying a translational ubq-1 construct consisting of 938 bp of ubq-1upstream sequences fused to IacZ (ubq93S-IacZ) expressed f3 -galactosidase in embryos and in atissue general manner in 20% of staining Li larvae. Somatic expression in later stages wasusually confined to body muscle. Progressively larger deletions extending from the 5’ end ofubq938-IacZ did not significantly alter the pattern of expression until 827 bp of sequence hadbeen removed. Thus sequences upstream of the transcriptional start site, including a GIG richblock and a sequence resembling a TATA box (GAATAA) are not required for ubq-1 expression.11AbstractMoreover, a basal level of expression was maintained in embryos when 903 bp had been deleted.These results suggest that some of the regulatory elements required for efficient expression ofubq-1 may reside within the transcribed region of the gene; alternatively they must lie morethan 1.7 kb upstream or 0.8 kb downstream of this region. PCR analysis indicates that RNAmolecules transcribed from the ubq938-IacZ and ubq 827-IacZ transgenes are trans-spliced toSL1, as is ubq—1 RNA.111Table of ContentsTABLE OF CONTENTSAbstract.iiTable of Contents IVList of Figures ixList of Tables XiList of Abbreviations XIIAcknowledgements XVDedication xviChapter I: Analysis of hspl6 Expression in C. elegans 1A. Introduction 11. General Introduction 12. Events in the Stressed Cell 23. Functions of the Heat Shock Proteins 33.1 HSP11O 33.2 HSP9O 43.3 HSP7O 53.4 HSP6O 73.5 Low MW HSPs 84. Heat Shock Proteins in Development, Differentiation and Growth 10 44.1 Developmental Control of the Heat Shock Response in Xenopus 104.2 Developmental Regulation in Drosophila 114.3 Developmental Regulation of hsp7O 125. The Role of Low MW HSPs in Thermotolerance 146. Regulation of the Heat Shock Response 156.1 Transcriptional Control in Eukaryotes 166.2 Post-transcriptional Regulation 19ivTable of Contents6.3 Translational Control .216.4 Recovery from Heat Shock 237. Caenorhabditis elegans as a model system 258. The Heat Shock Response of C. elegans 268.1 The Small HSPs of C. elegans 269. The Present Study 30B. Materials and Methods 311. Construction of hspl6-IacZ Fusions 312. Bacterial Transformations 323. Double Stranded Sequencing of IacZ Fusions 324. Maintenance of Strains 335. Establishment of Transgenic C. elegans Strains 346. Selection of Transformed Progeny 347. Viable Freezing of Transgenic Strains 358. Heat Shock Conditions 359. Identification of 3-gaIactosidase Staining Cells 3610. Preparation of Transgenic Genomic DNA 3611. Southern Transfer and Analysis of Transgenic DNAs 3712. DNA Dot Blot Procedures 3713. Labelling of Radioactive Probes 3814. Hybridization Conditions 39C. Results .401. Construction of hspl6-IacZ Fusions and Selection of Transformants 402. Expression of the hspl6-IacZ Transgene is Temperature Dependent 433. Establishing Standard Heat Shock Conditions for Experiments 464. The Transgenes 48.1 C and 41 .2C are Expressed in a Tissue General Manner46VTable of Contents5. Quantitative Differences in the Tissue Specific Expression of the hspl6 GenePairs 546. Elimination of One Promoter from an hspl6 Gene Pair Significantly ReducesSomatic Tissue Expression Without Affecting Embryonic Expression 607. Southern Analysis of Transgenic Strains 648. Determination of iacZ Transgene Copy Number 69D. Discussion 741. The IacZ Trangenes are Correctly Expressed in Response to Heat Shock invivo 742. Transgene Copy Number versus Expression Levels 743. Expression of the hspl6 Gene Pairs is Tissue General 754. Tissue Differences in the Expression of hspl6-IacZ Transgenes 765. Cooperative Interaction of HSEs May Enhance Expression 776. Developmental Regulation of the hspl6s in C. elegans 797. Conclusions 808. Future Prospects 81Chapter II: Analysis of Polyubiquitin Gene (ubq-1) Expression 84A. Introduction 841. General Introduction 842. Ubiquitin Mediated Proteolysis 843. Ubiquitin as a Regulatory Protein 864. Ubiquitin and Heat Shock 875. Ubiquitin and Chromatin 876. Ubiquitin at the Cell Surface 897. Ubiquitin and Myofibril Assembly 898. Ubiquitin Gene Structure 89viTable of Contents8.1 Polyubiquitin Gene Structure .898.2 Polyubiquitin Gene Expression 908.3 Ubiquitin Fusion Genes 938.4 Ubiquitin Like Genes 949. Trans-splicing 9410. The Present Study 95B. Methods 981. Maintenance of Strains 982. Construction of ubq-1-IacZ Fusions 983. Establishment of Transgenic Strains 984. Identification of f3-galactosidase Staining Cells 985. Heat Shock Conditions 996. Preparation of RNA 997. Preparation of cDNA 998. Polymerase Chain Reactions 1009. Separation and Analysis of PCR Products 10010. Southern Analysis of PCR Products 101C. Results 1021. Construction of ubq-1-!acZ Strains 1022. Expression of the ubq938-!acZ Transgene is Constitutive but ShowsDevelopmental Tissue Specificity 1043. Expression is not Diminished until 827 bp of Sequence has been Deletedfrom ubq938-IacZ 1074. The ubq938-IacZ and ubqA827-IacZ Transcripts are Trans-spliced 112D. Discussion 1161. Expression of ubq-1-IacZ Transgenes in Nematodes 116viiTable of Contents2. Ubq-1 Expression is not Significantly Heat Inducible .1183. Trans-splicing of ubq-1-IacZ Transcripts 1194. Analysis of the ubq-1 Promoter 1195. Conclusions 1216. Future Prospects 121References 123Appendix 136viiiList of FiguresLIST OF FIGURESFig. 1 Organization of the hspl6 loci of C. elegans 27Fig. 2 Construction of the hspl6-IacZ transgenes 41Fig. 3 Heat shock dependence of IacZ expression 44Fig. 4 Expression of f3-galactosidase is proportional to temperature in hspl6-IacZ transgenicanimals 47Fig. 5 LacZ expression of the 48.1C and 41.2C transgenes in response to heat is tissue general50Fig. 6 Cell types expressing hspl6-IacZ transgenes upon heat shock 52Fig. 7 Expression of hspl6-exonl fusions 55Fig. 8 Expression of the 48P transcriptional fusion in heat shocked transgenic animals 61Fig. 9 Southern analysis of Sma 1/Apa I digested genomic transgenic strain DNAs probed with nicktranslated pPD16.43 65Fig. 10 Southern analysis of transgenic genomic DNA probed with primer extended iacZ products67Fig. 11 Dot blot analysis of transgene copy number 70Fig. 12 The non-lysosomal ATP dependent proteolytic pathway 85Fig. 13 Organization, transcription and processing of the polyubiquitiri gene, ubq-1 of Caenorhabditiselegans 96Fig. 14 Construction of ubq-1-IacZ fusions 103Fig. 15 In situ staining of f-gaIactosidase activity in PC36 105Fig. 16 Deletion analysis of the ubq938-IacZ transgene 108Fig. 17 Expression of ubq-1-IacZ deletion transgenes 109ixList of FiguresFig. 18 Amplification of trans-spliced ubq-1-lacZ RNA 113Fig. 19 Testing the integrity of the PCR amplification scheme 114Fig. 20 Analysis of PCR products amplified from ubq-1-iacZ RNA 115xList of TablesLIST OF TABLESTable 1. Tissue distribution of 3-galactosidase staining in transgenic C. elegans heat shocked for15-120’ at 33°C 58Table 2. Distribution of J3-galactosidase activity in trangenic strains carrying transcriptionalfusions 63Table 3. Determination of IacZ trangene copy number in transgenic strains selected for extensiveanalysis 71Table 4. Comparison of expression patterns between 1 .48E1 strains 72Table 5. Comparison of polyubiquitin genes among several species 91xiList of AbbreviationsLIST OF ABBREVIATIONSAMP adenosine-5’-monophosphateATP adenosine-5’-triphosphatebp base pair(s)10 % Blotto 0.1 g/ml skim milk powder (Carnation); 0.2 % Na azideBSA bovine serum albuminC-terminus carboxy terminuscDNA DNA complementary to coding strandC. elegans Caenorhabditis eleganscpm counts per minuteDAPI 4’,6-diamidino-2-phenyl-indole1 % DenhardVs 0.01 g/ml of each of BSA, Ficoll, PVP, SDSDNA deoxyribonucleic aciddATP deoxyadenosine-5’-triphosphatedCTP deoxycytidine-5’-triphosphatedGTP deoxyg uanosi n e-5’-triphosph atedH2O distilled waterdTTP d eo xyth ym idi n e -5’-t rip h os ph atekIATP dideoxyadenosine-5’-triphosphatethDTP dideoxycytidine-5’-triphosphateckK3TP dideoxyguanosine-5’-triphosphateddTTP dideoxythymidine-5’-triphosphateEl ubiquitin activating enzymeE2 ubiquitin conjugating enzymeE3 ubiquitin ligaseE. coil Escherichia coilxiiList of AbbreviationsEDTA ethylenediamine tetraacetic acidHSC(s) heat shock cognate protein(s)hsc heat shock cognate protein geneHSP(s) heat shock protein(s)hsp heat shock protein geneHSF heat shock transcription factorHSE heat shock elementkb kilobase pair (s)kDa kilodaltonsMDa megadaltonsmRNA messenger RNAmHSP mitochondrial HSPMMLV Moloney murine leukemia virusM9 3 % KH2PO4; 6 % Na2HPO4; 5 % NaCI(w/v); 1 mM MgSO4N any base (i.e.adenine, cytidine, guanine, thymidine)N-terminus amino terminusNG\A nutrient growth mediant nucleotide(s)PCR polymerase chain reactionPP1 pyrophosphatePVP polyvinyl pyrolidineRCF relative centrifugal forceRNA ribonucleic acidRNase ribonucleaseS. cerevisiae Saccharomyces cerevisiae20 X SSC 3.6 M NaCI; 0.3 M Na Citrate, pH 7.0SDS sodium dodecyl sulphatexliiList of AbbreviationsSL splice leaderSSPE 180 mM NaCI; 1 mM EDTA; 10 mM NaH2PO4, pH7.4TBE 90 mM Tris-borate pH 8.3; 2 mM EDTATaq Thermus aquaticusTE 10 mM Tris-HCI, pH 8.0; 1 mM EDTATris Iris (hydroxymethyl) aminomethaneLa ubiquitin protein monomerXgal 5-bromo 4-chioro 3-indolylgalactosidasexivAcknowledgementsACKNOWLEDGEMENTSFirst and foremost, I would like to thank my supervisor Peter Candido for hisenthusiastic support during this endeavour. In particular, I am extremely grateful that he gaveme the opportunity and lattitude to pursue the project of my choice even though success was notguaranteed and the necessary equipment was not initially available. Four years and amicroinjection set-up later, I can honestly say that it was worth it. I would also like to thankthe other members of my supervisory committee- Don Moerman, Ross MacGillivray and TomGrigliatti-for their well-respected advice and counsel, lively committee meetings (the beer andcookies helped I’m sure) and for reading my thesis so quickly without complaining too much.To my collaborators, Don Jones and Dennis Dixon, thanks for sharing constructs, adviceand imagery from Thomas Hardy novels! To all the members of the Candido lab who have comeand gone during my internment at this institution (Don, Roger, Esther, Robert, Mike, Dave,Josie, Mei and Michel) a hearty thanks for enduring the spontaneous yelps of “kill, kill, killduring microinjection sessions, the pictures of blue worms plastered about the lab, the latenight opera singing and the “question of the day”.To all my friends from the Biochemistry department and mother programs, thanks forall the intriguing and boisterous conversations and for letting me “borrow” stuff. You’ve givenme a new perspective on life and science.I am grateful to Andrew Fire (Carnegie Institute in Baltimore) for donating nematodeexpression vectors to the cause and to Jim Preiss (Fred Hutchinson Institute in Seattle) for anexcellent tutorial on cellular identification in C. elegans.My warmest thanks to the members of the office staff for their patient printing of mythesis and papers, and for dealing on my behalf all too many times with irate financial personneletc.I would like to thank my family for their support in this venture. In particular myfather who as a Professor in Animal Science first introduced me to Genetics and my five elderbrothers for setting such “shining examples”. Last but not least, I would like to acknowledgethe loving support and numerous cups of tea my husband Peter has given me throughout thisdegree. Thank you for believing in me.xvDedicationDEDICATIONThis thesis is dedicated to the memory of my mother, Gabrielle Mary WellingtonStringham, who died of cancer on August 24, 1989 when this work was still in progress. Shewasn’t a scientist but she believed that every individual has an obligation to make this world abetter place for others. This is the ideal scientists must strive for. Let us collaborate- notcompete, let us remember who we work for, why we work and that we are but tiny players in amuch greater universe. Thank you Mum for making my world definitely a better place.xviChapter 1: IntroductionCHAPTER I: ANALYSIS OF HSP16 EXPRESSION IN C. ELEGANSA: INTRODUCTION1. General IntroductionThe heat shock or stress reponse is a universal phenomenon characterized by theinduction of a unique set of polypeptides called the heat shock proteins (HSPs) coupled with thesimultaneous repression of normal cellular protein synthesis. Golschmidt, in 1935, suggestedthat heat shock alters the pattern of gene expression: this was experimentally substantiatedwhen Ritossa documented that elevation in temperature from 25°C to 37°C induced theformation of specific chromatin puffs in the giant chromosomes of the salivary glands inDrosophila (Ritossa, 1 962). Later studies showed that these puffs represent sites of de novoRNA synthesis (Tissières et al., 1974) and the link between heat shock and gene regulation wasfirmly established.Heat shock in Drosophila induces the formation of eight distinct polypeptides (Moran etal., 1978) which range in mass from 22 to 83 kD. Generally, the HSPs can be subdivided intofamilies on the basis of their size as determined by migration in one or two dimensional gels: the110 kD HSPs, the 80-90 kD HSPs, the 7OkD HSPs, the low molecular weight (20-30 kD)HSPs, and ubiquitin (8kD). Ubiquitin is the subject of Chapter Il and will not be discussedfurther here. These proteins are highly conserved among both prokaryotic and eukaryoticspecies. For example human HSP7O shares 50 % sequence identity with HSP7O from E. coli,with the similarity reaching 96 % in some domains of the protein (Bardweil and Craig, 1984).This conservation suggests that HSPs have an essential protective role in all cells. However,even HSPs appear to be limited in their capacity to rescue a cell from thermal stress, sinceprolonged heat shock will eventually result in cell death.Since the discovery of the heat shock reponse, it has been shown that a variety of otheragents evoke similar changes in gene expression culminating in the repression of normal1Chapter 1: Introductionprotein synthesis and the induction of HSPs. These inducers include heavy metals, arsenite,ethanol, amino acid analogues, certain ionophores, mitochondrial inhibitors and viral infection(Ashburner and Bonner, 1979; Thomas et al., 1982). Thus the phenomenon is now moreaccurately referred to as the stress response, even though heat shock is the most frequentmethod employed by investigators to elicit the response.2. Events in the Stressed CellThe physiology and biochemistry of the cell undergo dramatic changes when confrontedby stress. Firstly, there is a cessation in cellular growth, the pH plummets, and there is anincrease in the levels of cytosolic calcium (Hammond et al., 1982; Findly et al., 1983). In thenucleus, protein and nucleic acid complexes become insoluble (Littlewood et al., 1987)resulting in aggregation of granular components (Welch and Suhan, 1985). In addition rod-like actin-containing filaments form and cross the nucleus (Welch and Suhan, 1985). Withinthe cytoplasm the proportion of actin filaments increases while the intermediate filamentnetwork collapses around the nucleus (Falkner et al., 1981). Mitochondria swell andintramembranous proteins denature (Lepock et al., 1983; Welch and Suhan, 1985).Ribosomes and mitochondria move to the perinuclear region and associate with theintermediate filaments (Welch and Suhan, 1986). The Golgi complex fragments into separatevesicles which become dispersed throughout the cytoplasm (Welch and Suhan, 1985).Heat shock also forces a change in energy metabolism. Stressed cells tend to shift fromaerobic to glycolytic pathways for energy production as the intracellular levels of ATP droprapidly due to mitochondrial dysfunction (Findly et al., 1983; Christiansen and Kvamme,1969; Mondovi et al., 1969; Dickson and Oswald, 1976). Lactic acid accumulates as aconsequence of this shift (Hammond et al., 1982).It is unclear which if any of these events actually triggers the stress response. In 1980Hightower proposed that HSPs are synthesized in response to the generation of abnormalproteins. He observed that addition of the amino acid analogue canavanine, induced the synthesis2Chapter I: Introductionof HSPs in chick embryo and mammalian cells (Hightower, 1980). Since then several lines ofevidence have given support to this argument. For example, Drosophila mutants whichsynthesize altered actin in the indirect flight muscles constitutively express HSPs in that tissue(Hiromi et al., 1986). In addition, mammalian and yeast cells which cannot degrade abnormaland short lived proteins, due to defects in enzymes of the ubiquitin mediated proteolyticpathway, constitutively express elevated levels of heat shock proteins (Finley et al., 1984;Ciechanover et al., 1984; Seufert and Jentsch, 1990; see Ch. II, section A4 for further detail).The effect of abnormal proteins in the stress response was directly studied bymonitoring the expression of a hsp7O-IacZ fusion in Xenopus oocytes which had been injectedwith native or denatured proteins. When bovine serum albumin (BSA) or f3-lactoglobulin wasinjected in the native form, expression of the hsp7O-lacZ fusion was not significantly altered.However, injection of denatured BSA or 3-lactoglobulin increased j-galactosidase activity 10fold (Ananthan et al., 1986). This suggests that the mere presence of abnormally foldedproteins can induce the stress response. Alternatively, it may not be the presence of abnormalproteins alone which elicits the response but rather the way those abnormalities affectdownstream metabolic and developmental pathways. Degradation of proteins for example couldhave enormous impact on energy metabolism by inactivating enzymes of aerobic metabolism(e.g. of the TCA cycle, oxidative phosphorylation etc.) and this may be the actual trigger of thestress response.3. FUNCTIONS OF THE HEATSHOCK PROTEINS3.1 HSP11OVery little is known about the function of this family of stress proteins. In mammals,these proteins are synthesized constitutively but are induced five fold upon heat shock. HSP11Oproteins localize to the nucleolus during both non-shock and heat shock conditions where theybind either directly to RNA molecules or form complexes with RNA binding proteins (Subjeck etal., 1983; Subjeck and Shyy, 1986). The yeast homologue HSP1O4 is a recently identified3Chapter 1: Introductionmember of the CIpAICIpB family of nucleotide binding proteins (Parsell et al., 1991). Sitedirected mutagenesis of the two putative nucleotide binding sites of HSP1O4 revealed that bothsites are essential for thermotolerance. The thermotolerant defect in HSP1O4 mutants can besuppressed by over-expression of HSP7O suggesting that the function of these two classes ofproteins is at least partially interchangeable (Parsell et at., 1991).3.2 HSP9OThis class of HSPs is present in abundance in the cytoplasm of normal cells. Upon heatshock, a small proportion of HSP9O translocates to the nucleus. In vertebrates, HSP9O seemsto regulate steroid hormone receptor action by masking the DNA binding site of the receptor inthe absence of hormone (Sanchez et a!., 1985; DeMarzo et at., 1991). HSP9O is highlyphosphorylated on seriries and threonines and is transiently associated with several kinases,including the tyrosine kinases encoded by oncogenes (Oppermann et al., 1981; Ziemiecki et al.,1986), and the eukaryotic translation-initiation factor kinase which modulates thephosphorylation and hence activity of 1F2. For this reason it has been hypothesized that HSP9Oin the normal cell regulates both transcription and translation in response to developmentalcues. During heat shock, it may be detrimental to proceed with regular developmentalprograms; signals may be faulty, delivered inappropriately or not at all, and execution ofpathways in response to those signals may be physically impossible. By inactivating hormonereceptors and kinases, HSP9O may prevent potentially deleterious mistakes in developmentaland growth programs.HSP9O also complexes non-covalently with actin and tubulin of the cytoskeleton and ithas been postulated that in this role, HSP9O acts as a chaperone to stabilise the cytoskeletalnetwork during heat shock (Schlesinger, 1990 ).4Chapter 1: Introduction3.3 HSP7OThe HSP7O family is the most highly conserved of the HSPs. Constitutively expressedmembers of this family are distributed throughout the cytoplasm and the nucleus in theunstressed cell. After heat shock however, newly synthesized HSP7O localizes to the nucleusand nucleolus (Welch and Feramisco, 1984). The current model regarding HSP7O functionsuggests that these proteins serve as molecular chaperones by binding to hydrophobic sites ofnascent or denatured polypeptides thus preventing the formation of inappropriately foldedinsoluble aggregates (Beckmann et al., 1990; reviewed in Lindquist and Craig, 1988;Schlesinger, 1990; Pelham, 1988). In the unstressed cell, HSP7O acts transiently to modulatethe unfolding, translocation and re-folding of newly synthesized proteins across membranes(Munro and Pelham, 1986; Chappell et al., 1986; Deshaies et al., 1988; reviewed in Peiham,1988). Functions in the heat stressed cell are thought to be similar but the target is shiftedfrom nascent polypeptides to thermally denatured structural proteins or inactivated enzymes(Gaitanaris et al., 1990; Skowyra et al., 1990; Beckmann et al., 1990; reviewed in Pelham,1988). HSP7O may keep these proteins soluble until they can be degraded by proteolyticpathways or in the case of salvageable proteins it may promote their re-folding.One HSP7O, the DnaK gene product, is an abundant protein in bacteria and is essentialfor cell division and growth (Bukau et al., 1989). This gene was originally isolated as amutation in bacteriophage growth. In particular, DnaK mutants were defective in lambda DNAreplication. Since then it has been shown that the DnaK gene product can refold and reactivateheat-inactivated RNA polymerase in an ATP dependent manner (Skowyra et al., 1990). This E.coil HSP7O shares approximately 48% amino acid sequence identity with eukaryotic HSP7Os(Bardwell and Craig, 1984).The Immunoglobulin heavy chain binding protein (BiP), previously, identified as glucoseregulated protein GRP78 of mammals, is a constitutive, non-inducible HSC (heat hock cognate)70 which facilitates the folding and assembly of secreted or membrane proteins which have beentransported across the endoplasmic reticular membrane (Munro and Pelham, 1986). A5Chapter I: Introductioncarboxyl terminal sequence, KDEL, distinguishes BiP from cytoplasmic HSP7Os and promotesits retention in the endoplasmic reticulum (ER) (Munro and Pelham, 1987).Another HSC7O has been identified as the uncoating ATPase which binds to and dissociatesthe clathrin light chain (Chappell et al., 1986).Recent studies suggest that cytosolic HSP72 and HSP73 bind co-translationally withnascent peptides because nascent polypeptides released prematurely from polysomes in vivowere found complexed with HSP72 and HSP73 (Beckmann et al., 1990). This provides directevidence that HSP7O in the unstressed cell chaperones the newly synthesized proteins. Instressed cells, association with HSP72 and HSP73 was prolonged and this effect was mimickedby the addition of L-azetidine 2-carboxylic acid , a proline analogue (Beckmann et al., 1990).These results suggest that it is not heat shock per se but the conformational state of the proteinwhich determines when dissociation of HSP72 and HSP73 occurs.In yeast, there are at least 9 members of the HSP7O family; three are localized tospecific cellular compartments and still others mediate translocation to these organelles. TheKar2 gene is a homologue of bip/grp78 and operates in the ER (Chirico et al., 1988). Anotherprotein, mHSP7O, encoded by the sscl gene, is an essential protein of the mitochondrial matrix(Deshaies et al., 1988) Temperature sensitive mutants of sscl fail to transport proteins intothe mitochondria. This HSP7O appears to function just within the mitochondrial membrane as itwas isolated in a complex with the membrane bound precursor polypeptide and an outermembrane protein (Deshaies et al., 1988). It is hypothesized that SSC1-HSP7O bindsimported precursors briefly before HSP6O interaction (See section 3.4) and actually pulls thepolypeptide through the membrane. As with all other HSP7Os, ATP hydrolysis drives release ofSSC1 from the protein (Deshaies et al., 1988).Thus the HSP7Os chaperone newly synthesized polypeptides to their final destinationswithin the cell. In the cytoplasm, HSP7Os unfold and bind to peptides as they are translated,and escort them to various cellular organelles. Within these compartments are other HSP7Oswhich may drive the translocation of the unfolded polypeptides through membranes into6Chapter 1: Introductioncellular compartments while still other HSP7Os and HSP6Os facilitate the appropriate refolding and assembly of the precursors within organelles.3.4 HSP6OThe HSP6O proteins possess “chaperonin like” functions similar to the HSP7O familyexcept that they mediate activities on the inner side of various cellular organelles ineukaryotes. Members of this family form complexes with polypeptides, have ATPase activity,and participate in the folding and assembly of proteins. These HSPs are constitutivelysynthesized in growing cells but are also mildly heat inducible.In bacteria, the groEL gene encodes a HSP6O which is essential for the assembly oflambda phage and for growth. The demonstration that GROEL protein participates in the foldingof pre-f3-lactamase also suggests that HSP6Os may play a role in protein secretion (Kusukawaet al., 1989). The ATPase activity of the groEL gene product is modulated by interaction withthe groES gene product (Tilly and Georgopoulos, 1982; Chandrasekhar et al., 1986). Recentexperiments demonstrated that GROEL could promote the assembly of RuBisCO (ribulose-1 ,5-biphosphate carboxylase-oxygenase) when its protein subunits were expressed from a plasmidtransformed into E. coil (Goloubinoff et al., 1989).Conditional mutants in the yeast groEL homologue, mif-4 are phenotypically defective inmitochondrial function; they can still import proteins to the mitochondria at restrictivetemperatures but cannot assemble them properly once they are inside the organelle. Analysisof the mif-4 gene product has revealed that it is a homo-oligomer of 14 subunits which isessential for re-folding of imported mitochondrial matrix proteins (Reading et al., 1989).In the chloroplasts of plants, the GROEL eukaryotic homologue RuBisCO-binding proteinmediates the identical process, the assembly of the active RuBisCO enzyme complex(Hemmingsen et al., 1988). Thus the functions of the HSP6O family are highly conservedbetween prokaryotes and eukaryotes.7Chapter 1: Introduction3.5 LowMWHSPsIn comparison to the other HSPs, the function(s) of the small HSPs are not understood.These proteins have been isolated from plants, yeast, invertebrates, and vertebrates and rangein size from 15 to 30 kD. All small HSPs contain a carboxyl terminal domain which showsextensive homology to the cx-crystallins of the vertebrate eye lens, suggesting that the acrystallins were derived from an ancestral heat shock protein which possessed propertiessuitable for lens function. These proteins form large stable aggregates in the lens whichprovide the transparency of the lens for its lifetime.All of the small HSPs isolated to date form massive aggregates in vivo. In humans, thereis only one small HSP, HSP28, which is constitutively synthesized in small amounts. In theunstressed cell, HSP28 forms soluble aggregates of 200-800 kDa in the cytosol. Upon heatshock, HSP28 migrates to the nucleus and forms even larger, up to 2 MDa, relatively insolublehomomeric complexes (Arrigo et al., 1988). During recovery there is a return to the pre-heatshock state, and HSP28 redistributes in the soluble cytoplasmic fraction (Arrigo et al., 1988).The localization of small heat shock proteins upon heat shock does not appear to beuniversal however. In chicken embryo fibroblasts, high molecular weight insoluble cytoplasmicaggregates develop from soluble particles after a second heat shock and do not re-distribute tothe nucleus (Collier et al., 1988).Tomato small HSPs also form high molecular weight insoluble cytoplasmic granules uponheat shock, and these granules have been found to be associated with mRNAs, specifically thosewhich are not translated during heat shock (Nover et al., 1989). This suggests that onefunction of the small HSPs may be to protect mRNAs during stress so that they are immediatelyavailable for translation upon recovery. Alternatively, the small HSPs may sequester mRNAsduring stress so that they are not translated under adverse conditions or to free up thetranslation machinery for HSP production.8Chapter 1: IntroductionThere are also situations in which small HSPs migrate to specific organelles upon heatshock: HSP3O of Neurospora for example relocates to mitochondria upon heat shock(Plesofsky-Vig and Brambl, 1990).HSP27 of Drosophila forms large insoluble aggregates upon heat shock which arelocalized to the nucleus. HSP27 is also hormonally regulated in the absence of stress (SeeSection 4.2). While ecdysterone induced HSP27 also localizes in the nucleus, the aggregates aresoluble (Beaulieu et al., 1989). In Drosophila there are four small HSPs raising thepossibility that the aggregates which form after heat shock may actually be heteromeric.In a quest to determine the function of small HSPS, McGarry and Lindquist transformedflies with an antisense hsp26 construct. While hsp26 expression was not eliminated by thisstrategy, HSP26 synthesis in heat shocked flies was dramatically reduced. HSP26 repressiondid not affect induction of other HSPs, nor was recovery and the resumption of normal cellularprotein synthesis delayed (McGarry and Lindquist, 1986). The implication here is that HSP26does not play a major role in the regulation of the heat shock response.Recent structural analysis of purified mouse HSP25 showed that it is organized intohigh molecular weight complexes of about 730 kDa or equivalent to 32 HSP25 monomers(Behlke et al., 1991). The authors proposed an arrangement of hexagonal packing to achievethe 15-18 nm in diameter spherical structure observed using electron microscopy (Behlke etal., 1991). In addition, they demonstrated that on two dimensional gels there are at least threeisoforms of HSP25, two of which are phosphorylated (Behlke et al., 1991). Another studyfrom the same laboratory isolated and sequenced phosphopeptides from HSP25 and determinedthat phosphorylation occurred specifically at two serine residues present in a kinaserecognition region which is conserved amongst small mammalian HSPs (Gaestel et al., 1991).Phosphorylation was not required to assemble HSP25 into large aggregates however (Behlke etal., 1991). The cL-crystallins have also been shown to be phosphorylated in vivo (Spector etal.,1985) and small HSP phosphorylation has been correlated with the acquisition ofthermotolerance (See section 5). Thus while phosphorylation does not appear to berequired for9Chapter I: Introductioncomplex assembly, it presumably is required for the function of these complexes during heatshock and in the vertebrate eye lens.4. Heat shock proteins in development, differentiation, and growthMany heat shock proteins can be induced by other agents which presumably exertdeleterious effects on the cell. However, some HSPs (e.g. HSP7O) have relatives (e.g. HSC7O)which are constitutively expressed in the absence of heat shock. This implies that heat shockproteins perform functions which are required by a normally growing cell as well as a stressedcell. While many of these HSC proteins show temporal and spatial expression patterns, thedevelopmental induction of the small hsps which do not have family members that areconstitutively expressed is of particular interest. Moreover, inducibility of the heat shockresponse itself and the acquisition of thermotolerance seem to be under developmentalregulation. The following examination of these processes provides some examples of the ever•growing documentation of HSP induction in the absence of stress.4.1 Developmental control of the heat shock response in XenopusDevelopment of the heat shock response in the frog is a complex process. Prior toovulation, in response to heat shock, oocytes translate HSPs 83, 76, 70, and 57 from preexisting maternal messages (Bienz and Gurdon, 1982; Browder et al., 1987). After ovulation,the small HSPs are also heat inducible (Browder et al., 1987). Fertilization dramaticallyterminates the heat shock response, and no heat shock proteins can be induced. However,constitutive HSP7O synthesis is derived from oogenic hsp7O RNA which has been retained(Browder et al., 1987). Extinction of the heat shock response in early embryos coincides witha period of extreme thermolability. The mid-blastula transition re-introduces the heat shockresponse when transcription of the embryonic genome commences (Bienz, 1984). The firstHSPs which are inducible at this stage are HSP68, HSP7O, and HSP87. This period alsocorresponds to the acquisition of thermotolerance in embryos. By neurulation, the small HSPs10Chapter 1: Introductionand the HSP6O family as well as other members of the HSP7O family are inducible. Thus,inducibility of HSPs is acquired asynchronously (Browder et at., 1987; Nickells and Browder,1988).Not only is inducibility of each HSP acquired independently during development, butthere is also differential expression with regards to spatial location of synthesis, temperatureof maximum induction, and duration of synthesis during stress. For example, HSP35 isinitially inducible at the blastula/gastrula transition stage at temperatures above 35°C andusually only for the first 40 minutes of heat shock (Nickells et at., 1989). Inducibility of thisHSP drops considerably in late gastrulation, followed by a resurgence at neurulation (Nickellset at., 1989). HSP 35 is actually the glycolytic enzyme glyceraldehyde-3-phosphatedehydrogenase (GAPDH) and the inducibility of this enzyme is inversely proportional to levelsof constitutive GAPDH (Nickells and Browder, 1988) suggesting that high levels of constitutiveGAPDH can adequately cope with the immediate crisis presented by heat shock. Thus theinducibility of particular HSPs at certain times or places in development may reflect themetabolic state of the tissue at that point in time.The adult frog continues to show differential synthesis of heat shock proteins in responseto stress. While undifferentiated erythroblasts can synthesize a variety of HSPs, matureerythrocytes synthesize only HSP7O (Winning and Browder, 1988).4.2 Developmental regulation in DrosophilaEvidence in Drosophila also indicates that hsp induction is absent in early development.and acquired at later stages. Oocytes constitutively express hsps 83, 28/27 and 26, while hsps70, 23 and 22 are neither expressed nor heat inducible (Zimmerman et at., 1983).The small hsps of Drosophila (Hsps22, 23, 26, and 27) are clustered within a 12 kb regionof the genome identified as locus 67B (Petersen et at., 1979; Corces et at., 1980; Voellmy etal., 1981). In the absence of stress, expression of the small HSPs occurs in imaginal discs andduring pupation and is controlled by the moulting hormone ecdysterone (Ireland and Berger,11Chapter 1: Introduction1982). Deletion analysis and transient chioramphenicol acetyltransferase (CAT) assays inDrosophila tissue culture cells showed that ecdysterone induction was mediated throughmultiple elements between -579 and -455 of the hsp27 promoter while heat induction wasregulated through three HSEs present between -370 and -270 (Riddihough and Peiham,1986). Dnase I hypersensitive sites centred at -522 and -293 of the hsp27 gene support thepositioning of these elements (Costlow and LIs, 1984). Surprisingly, similar studies of thehsp23 gene implicated sequences dissimilar to the elements defined for hsp27. In addition,experiments where hsp27 gene fusions were introduced into flies by P-element mediatedtransformation yielded slightly different results again (Hoffman and Corces, 1986; Hoffman etal., 1987). These discrepancies may reflect the inherent problems associated with transientassays of developmentally regulated genes in cultured cells. Since there is an established tissuespecificity for some of these genes (e.g. hsp2ô is expressed in the male germ line and imaginaldiscs (Glaser et al. 1986)) it is quite possible that data from tissue culture cells does notaccurately reflect the in vivo situation. Nonetheless, all of the experiments suggested that theelements regulating developmental expression of the hsps by ecdysterone are separate anddistinct from the HSEs.4.3 Developmental regulation of hsp 70There are many cases in which members of the HSP7O/HSC7O family are synthesized inresponse to developmental cues. Fetal rabbits at 25 days gestation are transcriptionally heatinducible for hsp7o while mRNA for its cognate is expressed constitutively in unstressed fetaltissues (Brown et al., 1985). Primitive cells of the avian erythroid lineage express HSP7Oconstitutively but are not heat responsive at either the transcriptional or translational levelwhile cells of the definitive lineage synthesize HSP7O after heat stress through a translationalcontrol mechanism (Banerji et al., 1987). Differential synthesis of HSPs also coincides withdifferentiation of the mammalian testis. In the immature mouse testis, HSP72 and HSP73 aresynthesized at low levels constitutively in the absence of stress, and HSP72 is strongly heat12Chapter I: Introductioninducible. The mature testis synthesizes HSP72 at slightly elevated levels after heat stress viatranslational control mechanisms. In addition, a novel isoform of HSP73 was detected inunstressed meiotic and post meiotic germ cells of the adult testis. Expression of this proteinwas not increased upon heat shock (Zakeri et aL, 1990). Thus, only cells of the somatic gonadwere heat responsive.Levels of HSP7O may be even more tightly controlled during growth. Synthesis ofHSP7O incresases during the late G1 to S phase, and peaks at G2 (Van Dongen et al., 1986;Simon et al., 1987). Other studies have shown that HSC7O proteins are under cell cycle controlvia an E1A like activity (Nevins, 1982; Kao et al., 1985).What exactly is the role of heat shock proteins in development? Some investigatorssuggest that expression of HSPs during development is an essential pre-requisite to thedevelopment of natural thermo- resistance or as it is more commonly referred to, intrinsicthermotolerance. Usually, the acquisition of thermotolerance correlates well with theacquisition of HSP inducibility. For example, decreased risk of lethality in mouse embryos hasbeen shown to be associated with the ability to synthesize inducible HSP68 (Muller et al.,1 985). But this does not explain the temporal and spatial expression patterns observed in theabsence of heat shock. Developmental programs and growth demand rapid changes in thearchitecture and metabolism of the organism. Not only must cells multiply but they mustbecome specialized, migrate, and form adhesions with other cells to create functional,communicating organs. Many cells are doomed to death and their corpses must be scavenged.Thus, the disintegration, re-distribution and ordering of multiplying cellular componentsduring devlopment may mimic the consequences of heat shock and the hormonally inducedexpression of hsps might meet the challenge imposed in specific situations. Alternatively, HSPsynthesis may be required to direct initial changes in gene expression during development.Studies in trypanosomes have indicated that when these parasites are transmitted from theirinsect vectors to their mammalian hosts they undergo heat shock by virtue of an increase from25°C to 37°C. This temperature shift induces differentiation from the promastigote into the13Chapter I: Introductionamastigote form and can be reproduced in vitro (Van der Ploeg et al., 1985). Thus heat shockmay act as a switch in determining developmental pathways.5. The role of low molecular weight HSPS in thermotoleranceThe concept of thermotolerance was conceived when it was noticed that a pre-shocktreatment could render an organism more resistant to a subsequent challenging stress whichwould otherwise be lethal (Milkman, 1966). That HSP production might play a role inthermotolerance was suggested by Mitchell et al. (1979) who found that mild heat treatmentswhich induced the synthesis of HSPs provided protection from heat induced lethality, anddevelopmental abnormalities in Drosophila. In addition, agents other than heat such as arsenite,ethanol, and re-oxygenation after hypoxia can induce thermotolerance (Li and Werb, 1982).In one study, Nguyen et al. (1989) followed the kinetics of reporter enzymessynthesized constitutively in mouse and Drosophila cells under the control of the SV4Opromoter when the cells were subjected to heat shock. Both f-galactosidase and luciferaserapidly became insoluble in both cell types. However, enzyme inactivation during heat shock orethanol induction was significantly reduced in pre-heat treated cells even when the challengingstress was applied 20 hours after recovery was initiated from the priming treatment. Since themouse does not produce endogenous -galactosidase or luciferase, the effect of thermotolerancecould not be attributed to any specifically designed mouse cell functions (Nguyen et a!., 1989).Several lines of evidence suggest that the small HSPs may be involved in the acquisitionof thermotolerance. Ecdysterone induction of the small HSPs in Drosophila tissue culture cellshas been shown to confer a thermotolerant phenotype, and Drosophila pupae are naturally moreheat resistant than other fly stages (Mitchell et al., 1979; Berger and Woodward, 1983). Athermo-intolerant Dictyostelium mutant is totally deficient in small HSP synthesis butotherwise synthesizes the complete set of HSPs upon heat shock. However, the levels of HSP7Oafter heat shock are somewhat reduced compared to wild type levels (Loomis and Wheeler,1982), and it may be this aspect which results in the thermosensitivity of the mutant.14Chapter 1: IntroductionHeat resistant mutants isolated from Chinese hamster lung cells have increased levels ofHSP27, particularly of the phosphorylated isomers of HSP27 (Chrétien and Landry, 1 988).When the protein synthesis inhibitor cycloheximide was added to normal Chinese hamster cells,HSP27 phosphorylation accompanied by thermotolerance was induced, without de nova proteinsynthesis (Landry et al., 1989). Thus phosphorylation of the small HSPs may be an importantcomponent in the acquisition of thermotolerance.6. REGULATION OF THE HEATSHOCKRESPcWSEThe rapid initiation of heat shock protein synthesis and the repression of normalcellular protein synthesis is controlled by a host of regulatory mechanisms. The followingdiscussion is drawn largely from data in eukaryotes, but particularly with regards totranscriptional regulation, the principles can often be transferred to prokaryotes.Prokaryotic heat shock genes form a regulon which is transcriptionally induced uponheat shock by the activity of a positive trans-acting factor. This factor is a specialized versionof RNA polymerase, a32 in E. coil, which acts through recognition of a cis-acting elementpresent upstream of heat shock genes. Increasing the rates of a32 synthesis by transformingbacteria with a plasmid containing the a32 gene, rpoH, under control of the lac promoterresulted in increased production of HSPs in the absence of stress indicating that levels of theprokaryotic transcription factor directly determines the extent of the heat shock response(Grossman et al., 1987).In eukaryotes, heat shock factor is a separate entity from RNA polymerase II, but it isthe action of this factor in concert with RNA polymerase II and other proteins (e.g. TATA factor)which triggers the enormous surge in hsp gene expression. In addition, eukaryotes employpost-transcriptional and translational control mechanisms to fine tune the response.15Chapter 1: Introduction6.1 Transcriptional Control in EukaryotesThe dramatic synthesis of heat shock proteins upon elevation in temperature is largelydue to de novo or increased transcription of heat shock genes. In eukaryotes this is mediatedthrough the binding of a multimeric positive trans-acting factor, heat shock factor (HSF) tocis-acting sequences, the heat shock elements (HSEs), which are usually located 80-150 bpupstream of the transcriptional start site. Originally defined as a 14 bp consensus sequence ofCTNGAANNTTCNAG (Pelham, 1982; Pelham and Bienz, 1982) a heat shock element is nowbelieved to consist of at least two or three modules of the sequence NGAAN, which can be presentin either orientation (Perisic et al., 1989).Many heat shock genes such as hsp7o of Drosophila, require elements other than theHSEs such as CCAAT boxes for optimal expression (Dudler and Travers, 1984). Others areexpressed at specific times during normal development in the absence of heat shock throughhormone binding elements which are distinct from the HSEs. (Previously discussed in section4.2).Recent work has focussed on the interaction between HSF and HSEs. HSF has beenpurified from yeast (Sorger and Pelham, 1987), Drosophila (Wu et al. 1987), and human(Goldenberg et al., 1988) and somewhat surprisingly, these proteins display a wide degree ofdivergence. Except for sequence identity in the DNA binding and the trimerization domains theyappear to be totally unrelated. This is in sharp contrast to the high degree of conservationobserved amongst heat shock proteins and the HSEs between species, and implies that there canbe many functional variations on the theme of heat shock factor. In yeast, the gene for HSF hasbeen recently cloned and characterized (Wiederrecht et al., 1988; Sorger and Pelham, 1988).It is a single copy gene which is essential for normal growth (Wiederrecht et al., 1 988;Sorgerand Pelham, 1988). Yeast must therefore require transcription of heat shock genesconstitutively or other unidentifed genes are expressed via HSF. Heat shock factor was isolatedfrom non-shocked yeast by affinity chromotography in which concatenated oligomers of HSEswere linked to a Sepharose column (Sorger and Pelham, 1987). Purified heat shock factor16Chapter 1: Introductionfrom unshocked yeast is relatively large (l5OkD); this is somewhat greater than the 93kD sizepredicted by the sequenced gene. Deletion analysis determined that the DNA binding domainresided within a basic region in the amino terminal portion of the protein. Originally, noobvious secondary structures commonly observed in DNA binding proteins were recognized inyeast HSF (e.g. zinc fingers, leucine zippers etc.) (Wiederrecht et al., 1988; Sorger andPelham, 1988). However, recent evidence suggests that the DNA binding domain of HSFpossesses a helix turn helix structure while the domain for oligomerization resembles a leucinezipper (Cbs et al. 1990; Rabindran et al. 1991; Schuetz et al. 1991).In yeast, HSF binds to HSEs constitutively and becomes transcriptionally activated afterheat shock (Sorger et al., 1987). In fact, levels of heat shock factor bound DNA isolated fromcontrol and heat shock cells are virtually equivalent. However, Sorger et al. (1987) noticedthat the mobility of HSF-HSE complexes on polyacrylamide gels shifted after heat shocktreatment and that at least five modified species were present. Addition of phosphatase reducedthis shift in mobility. In addition, they observed that the mobility shift increased linearly withtemperature increase. However, phosphorylation did not alter the affinity of HSF for HSE DNA.They suggested a model whereby in yeast, HSF binds to DNA under normal conditions. Heat shockpromotes phosphorylation of HSF at multiple sites which alters the conformation of HSF suchthat it can now complex or interact with other constituents of the transcriptional machinerysuch as TATA factor and/or. RNA polymerase to stimulate transcription (Sorger et al., 1987;Sorger and Pelham, 1988).In Drosophila and human cells, HSF binds to HSEs after heat shock (Wu et al., 1987;Sorger et al., 1987; Kingston et al., 1987; Zimarino et al., 1987), and this change in bindingaffinity is accompanied by phosphorylation. Binding and activation of HSF to HSEs in Drosophiladoes not depend on de novo protein synthesis suggesting that post translational modificationsalone may mediate conformational changes in HSF which allow it to interact with DNA. WhyDrosophila and human cells bind HSF after heat shock as opposed to the constitutive DNAinteraction in yeast is a puzzling question.17Chapter 1: IntroductionExperiments in yeast suggest that the transcriptional activity of HSF can actually beseparated into two distinct domains; one is required for sustained transcription at normalgrowth and moderate heat shock temperatures (i.e. 15-33°C) while another transientlyactivates transcription during a severe heat shock (i.e. >34.5° C) (Nieto-Sotelo et al., 1990;Sorger, 1990). Recently in Drosophila , using an anti-HSF antibody, Westwood et al. (1991)showed that during heat shock HSF localized specifically to heat shock puffs in polytenechromosomes as well as to sites of hormonally induced puffs which regress during heat shock.These data suggest that HSF may also act as a repressor of normal gene activity during heatstress (Westwood et al., 1991).While yeast possesses only one HSF gene, the recent discovery that there are at least twoHSF genes in humans and three in tomato raises intriguing possibilities (Schuetz et al., 1991;Rabindran et al., 1991; Scharf et al., 1990). Multiple HSFs may allow for the specialization ofthese proteins to perform different functions. Depending on the physiological conditions of thecell, a different HSF may be required to bind to the promoter to elicit a transient or sustainedresponse. This might explain why HSF binds to DNA after the stress has been implemented inanimal cells.Additionally, protein-DNA cross linking experiments by Gilmour and Lis (1986) andnuclear run-on transcription assays by Rougvie and Lis (1988) determined that the Drosophilahsp7O gene is transcriptionally engaged in vivo under non-shock conditions, and that RNApolymerase II has begun the synthesis of a nascent transcript of about 25 nucleotides. Theauthors suggested that RNA polymerase is stalled at this point and is unable to elongate thetranscript until heat shock activates its release. This would allow for rapid expression of heatshock genes in response to thermal stress by circumventing a potential delay in transcriptiondue to selection and binding of a specific HSF by already initializing the task. Other Drosophilagenes such as hsp2ô have also been shown to be transcriptionally engaged (Rougvie and Lis,1988;Thomas and Elgin, 1988). How widespread a phenomenon this characteristic is remainsto be seen.18Chapter 1: Introduction6.2 Post transcriptional regulationOne of the first indications that heat shock interferes with processing of transcripts wassupplied by the study of Mayrand and Pederson (1983) in which they found that heat shockdisrupts RNP particles in Drosophila. More recently, it has been demonstrated that heat shockdramatically alters the structure of the spliceosome itself. Immunological studies in bothDrosophila and mammalian cells showed that anti-snRNP antibodies, which recognize UisnRNPs, reacted similarly with preparations from both non-heat shocked and heat shockedcells. But while non-heat shocked cells reacted with anti-Sm antibodies, which detectcomponents of the Ui, U2, U4, and U5 snRNPs, heat shocked cell preparations were unable todo so (Welch and Mizzen, 1988). Cells which had been treated with a mild heat shock prior tothe challenging heat shock maintained Sm antigenicity (Wright-Sandor et al. 1989). Thus, heatpre-treatment appears to protect components of the splicing machinery.Accumulating evidence suggests that processing of pre-mRNA is much more sensitive tothe effects of heat shock than transcription. The Drosophila hsp83 gene, which contains a single5’ intron, is expressed at normal temperatures and is strongly induced by moderate heatshocks. However, poor expresssion was observed at high temperature heat shocks due to ablock in splicing of the pre-mRNA transcript. In contrast, transcription of the gene persisted athigh levels (Yost and Lindquist, 1986) resulting in the accumulation of large amounts ofunprocessed full length transcripts. Processing of an adh gene intron put under control of thehsp7O promoter was also inhibited by heat shock suggesting that the block in splicing isgeneral (Yost and Lindquist, 1986). This block in splicing persisted for two hours afterrecovery conditions were initiated. Heat shock has also been shown to disrupt cis-splicing inchicken, Dictyostelium, mammalian cells (Bond and Schlesinger, 1986; Kay et al., 1987;Maniak and Nellen, 1988; Bond, 1988) and trans-splicing in trypanosomes (Muhich andBoothroyd (1988). Kay et al. (1987) demonstrated that while transcription of a transfectednematode small hspl6 gene pair was normal in heat shocked mouse cells, splicing of introns19Chapter I: Introductionwas defective and unprocessed transcripts accumulated in the nucleus. During recovery thesetranscripts were processed and transported to the cytoplasm. In contrast, processing of thewild type hspl6 gene pair was normal in heat shocked nematodes and did not result in theaccumulation of unspliced transcripts. This suggests that the splicing apparatus may be lesssensitive to heat stress in C. elegans than in other organisms. Thus, heat shock genes in C.elegans may be able to afford the luxury of introns which hsps in other animals cannot.Yost and Lindquist (1986) demonstrated that pre-treatment of Drosophila cells with amild heat shock could rescue splicing of the hsp83 transcript during a severe heat shock whichwould otherwise have inhibited processing. When the translation inhibitor cycloheximide wasadministered with the heat pre-treatment before HSPs could be produced, splicing was laterdisrupted at high temperatures. Addition of cycloheximide after HSP synthesis had no effect onsplicing at subsequent high temperature heat shocks (Yost and Lindquist, 1986). Collectively,the above experiments suggest that HSPs themselves play a role in the protection and/ orrepair of the splicing apparatus during heat shock, either by stabilizing the components of thespliceosome or perhaps by participating directly in the process of splicing itself. Although it isnot clear which HSPs specifically might have these functions, the HSP7O family is a likelycandidate since its members consistently move to the nucleus during heat shock.In most species, the canonical heat shock genes are devoid of intervening sequences;hsp7O for example does not contain introns while the hsc7O genes do (Ingolia and Craig, 1982).Clearly this may be simply an adaptation to avoid the shortcomings of the splicing apparatusduring heat stress. Hsp83 is the only Drosophila heat shock gene which possesses an intron. Ithas been suggested that since this gene is already expressed at reasonably high levels duringnormal growth, there may be enough processed transcripts available to meet the increaseddemand for HSP83 protein during heat stress and that translational control might operate morein the regulation of HSP83 synthesis.20Chapter I: Introduction6.3 Translational ControlDuring heat shock, the synthesis of normal cellular proteins is actually repressed infavour of heat shock protein synthesis. Thus, induction of heat shock proteins alone is not thesole change in cellular activity integral to the stress response. While transcription andprocessing of normal cellular RNAs (unlike hsp RNAs) are severely hampered at elevatedtemperatures, levels of pre-existing RNAs are quantitatively maintained (Storti et al., 1980;Findly and Pederson, 1981; Lindquist, 1981; Petersen and Mitchell, 1981). Moreover, whennormal cellular RNAs are extracted from heat shocked cells they can be translated in cell freesystems in vitro. Thus the integrity of the RNA.transcripts themselves must also be maintained.Repression of normal protein synthesis is remarkably swift and precedes the appearanceof hsp RNA transcripts (McKenzie et al., 1975; Lindquist, 1981). Addition of actinomycin Dprior to heat. shock to prevent hsp transcription still results in repression of normal cellularprotein synthesis (Lindquist, 1981). Thus repression of normal protein synthesis andinduction of hsp gene expression can be uncoupled.Early studies by McKenzie et al. (1975) indicated that polysomes disappeared with therepression of normal protein synthesis. Subsequently, polysomes re-associated with newlytranscribed hsp mRNA5 (Lindquist, 1980). In another study, Drosophila lysates purified fromheat shocked cells preferentially translated hsp RNA while lysates isolated from cellsmaintained at 25°C showed no preference when incubated at the same temperature. Thus, oncespecificity was established it remained a property of the lysate. Mixing the lysates resulted intranslational activity intermediate between the two (Scott and Pardue, 1981; Lindquist,1981). When the heat shock lysate was supplemented with a fractionated crude ribosomalpellet from non-shocked cells translation of normal cellular RNAs ensued. Addition of theribosomal pellet from heat shocked cells to the 25°C lysate did not alter the translationalspecificity of the control lysate (Scott and Pardue, 1981). This suggests that a factor isolatedwith the ribosomal pellet is inactivated in heat shocked cells.21Chapter I: IntroductionRecent experiments by Marota and Sierra (1988) have shown that in heat shock lysates,translation of hsp RNAs is unaffected by the addition of cap analogues while translation of 25°CRNAs is further repressed. These authors suggest that normal cellular RNAs require capping tobe translated while heat shock RNAs by virtue of some property in their secondary structuremay be able to bypass this necessity. They suggest that it is the inactivation of cap bindingfactor which results in repression of normal cellular RNA translation (Marota and Sierra,1988; Marota and Sierra, 1989). However, heat inducible RNAs do appear to possess normalcap structures in vivo (Yost et al. 1990). Thus, while they may be able to bypass therequirement for capping in certain situations, it is unlikely that this alone could explain thehighly efficient preferential translation of heat shock RNAs.Rather, mutational analysis of the hsp7O and hsp22 genes of Drosophila has revealedthat sequences present in the 5’ untranslated leader contribute significantly to the ability of hspRNA to escape the block in translation conferred by heat stress. Firstly, when the 5’ end ofhsp7O including the untranslated leader was fused to other non-hsp sequences, the hybridtranscripts were efficiently translated at elevated temperatures (DiNocera and Dawid, 1983;Bonner et al., 1984). In another study, transcripts derived from hsp7O constructs containinginternal deletions in coding sequence were translated as efficiently as transcripts derived fromthe wild type gene. Deletion of the 5’ untranslated leader, in contrast, destroyed translatabilitywhile transcription of the fusion was unaffected (McGarry. and Lindquist, 1985). Theseexperiments suggested that the untranslated leader was important for translatability but did notindicate whether this property was confined to specific sequences in the leader or resulted fromproperties of its secondary structure.Leaders of heat shock genes in Drosophila tend to be fairly long, usually about 250 bp.To determine if specific signals are contained within this region, Klemenz et al. (1985)constructed various fusions of the Drosophila hsp7O gene to the alcohol dehydrogenase (adh)gene and transformed them into flies. A fusion of the hsp7O promoter to the adh gene wastranscriptionally heat inducible but mRNAs were not efficiently translated. However insertion22Chapter I: Introductionof the first 95 nucleotides of the hsp7o transcript was sufficient to confer preferentialtranslation of fusion transcripts at heat shock temperatures (Klemenz et al., 1985). Deletionanalysis of the hsp22 leader yielded similar results; 86% of the leader could be removed sinceonly the first 26 nucleotides were required for translation. Collectively, these data imply thathigh levels of transcription at elevated temperatures are not sufficient to achieve translation,and that sequences in the 5’ end of the leader are important for selective translation oftranscripts. Similar experiments have demonstrated that large deletions (+37 to +205 of the249 bp hsp7o leader, and +27 to +242 of the hsp22 leader) do not affect translatability(McGarry and Lindquist, 1985; Hultmark et al., 1986). On the other hand insertion ofadditional sequences near the 5’ end may, but not necessarily hamper translation. In onesituation, insertion of 39 nucleotides of sequences upstream of the transcriptional start site, atposition +2 of the hsp7O leader eliminated translation, while insertion of a duplication of thefirst 37 nucleotides of the leader at the same position did not diminish translation (McGarry andLindquist, 1985). Collectively, these data indicate that while conserved sequences at the 5’ endof hsp untranslated leaders are important, the structure of the leaders themselves may beequally important in signalling preferential translation of hsp transcripts.• There are also indications that signals in the 3’ ends of hsp messages might regulatesynthesis of HSPs. Experiments with human tissue culture cells have demonstrated that thehsp7O message is more stable after heat shock and that this stability is mediated through the 3’untranslated leader (Theodorakis and Morimoto, 1987). V6.4 Recovery from heat shockRecovery from heat shock implies a return to the normal physiological state of the cellprior to the initiation of stress. It therefore must involve the inactivation or repression of heatshock protein synthesis and the resumption of normal cellular protein synthesis. But implicitin the return to normalcy is the assumption that “all must be well” within the cell and that thecell is adequately prepared to resume normal metabolic pathways. For this to be true, hsps23Chapter 1: introductionmust have successfully fulfilled their role as protectors and repairmen from the deleteriouseffects of stress. So what determines when a cell has truly recovered from heat stress?Generally, at least two criteria must be met for recovery to take place. First of all there mustbe a return to normal physiological growth temperatures or removal of the stress agent.Secondly, there are indications that minimal quantities of HSPs are required before theresponse is shut down (DiDomenico et al., 1 982a), although HSP synthesis may not be requiredat all for recovery from very mild or short heat shocks. This implies that HSPs auto-regulatethe extent and duration of the response.In Drosophila cells, specific levels of HSP7O are consistently required before recoveryfrom heat shock can begin (DiDomenico et al., 1982b). Inhibition of HSP7O synthesis byartificial production of hsp7O antisense RNA delays recovery (Melton et al., 1988) whileinhibition of HSP26 synthesis by antisense RNA does not (McGarry and Lindquist, 1986). Stillother evidence suggests that the cell autoregulates levels of HSP7O relative to the extent of thestress (Stone and Craig, 1990). These data imply that HSP7O might be the sensor which cuesthe other components of the response to shut down.While the induction of HSP synthesis is an amazingly rapid response, maximal usuallywithin minutes, down-regulation of the response occurs over a much longer period of time,usually several hours. HSP synthesis does not cease immediately but rather decreases graduallywhile translation of the pre-existing normal RNA5 which were sequestered during heat shockgradually resumes. Inactivation of hsp gene transcription and selective degradation of existinghsp mRNAs results in a steady decline in HSP synthesis (DiDomenico et al., 1982a).Transcriptional activation of hsp genes is a highly coordinated event such that HSP synthesisoccurs in unison. In contrast, hsp mRNAs are repressed asynchronously in Drosophila startingwith repression of hsp7O and ending with repression of hsp82 (DiDomenico et al., 1982b;Lindquist and DiDomenico, 1985). The fact that hsp7O is down-regulated first providesfurther support to the above argument that this HSP may be the real autoregulator of theresponse.24Chapter 1: IntroductionWhat mediates the selective degradation of heat shock messages? In the previous sectionit was mentioned that hsp7O mRNA becomes more stable after heat shock and that this propertywas contained within the 3’ untranslated region of the message. Messages transcribed fromfusions of the adh gene and the 3’ untranslated sequences of hsp7o show degradation kineticssimilar to wild type hsp7O transcripts (Petersen and Lindquist, 1989). The 3’ untranslatedregion of hsp7O possesses elements with identity to 3’ sequences of normally unstable messagessuch as that of c-myc and c-los. Not surprisingly, heat shock stabilizes c-myc and c-los RNAs(Sadis et al., 1988). Thus while it is not clear how heat shock actually stabilizes certaintranscripts, it is this process which is partially responsible for the induction of HSP7Osynthesis during heat shock. During normal cellular growth and recovery, hsp7O messages areunstable, thus maintaining a low level of available HSP7O in non-shocked cells.7. Caenorhabditis as a model systemCaenorhabditis elegans is a small (less than 1 mm in length) free living soil nematodepossessing many qualities which are advantageous to the researcher. The life cycle of thisanimal is short, three to four days at room temperature, and it can be simply maintained in thelaboratory on Petri plates containing nutrient growth media ‘spread with a layer of bacteria.Alternatively, large quantities of animals for biochemical analysis can be cultured in liquid inlarge bottles (Sulston and Brenner, 1974).The transparency of this animal and its relatively simple anatomy has allowed thecomplete delineation of every cell lineage (Sulston, 1976; Sulston and Horvitz, 1977; Sulstonet at., 1983; Albertson and Thomson, 1976), including cell migrations and programmed celldeaths and of the animal’s neurocircuitry (White et at., 1986).In addition, powerful genetic tools are available in C. elegans. Numerous mutants havebeen isolated and while the animal is a self fertilizing hermaphrodite, males arisespontaneously by non-disjunction at a frequency of about 1 in 500 when stocks are maintainedat 20°C, allowing for crosses. Each hermaphrodite can produce about 250 progeny. The genome25Chapter I: Introductionis compact, about 5 times the size of the yeast genome and two thirds the size of Drosophilas.Moreover, the genome contains a low proportion of short-period repetitive sequences(Emmons et al. 1980). The average intron in C. elegans is 50-100 bp in length. As a result ofthe efforts of the MRC laboratory in Cambridge and the laboratory of R. Waterston in St. Louis,most of the genome has now been cloned and ordered as a series of cosmid and yeast artificialchromosome (YAC) clones (Coulson et al., 1986; Coulson et al., 1988). Sequencing of thegenome of C. elegans has begun.Finally, of particular relevance to the present study, transformation, both integrativeand extrachromosomal, has become an efficient and relatively simple process by which reversegenetics can be employed to study gene expression and cell biology (Fire, 1986).8. The Heat Shock Response of C. elegansElevation of the temperature from 25°C to 31°C - 34°C induces the synthesis of HSPscorresponding to the HSP9O, HSP7O and small HSP families characteristic of other species(Snutch and Baillie, 1983). But while C. elegans synthesizes all families of HSPs, research hasfocussed on the HSP7O family (Snutch and Baillie, 1984; Snutch et al., 1988) and the small 16kDa HSPs (Russnak and Candido, 1985; Jones et al., 1986).8.1 The small hsps of C. elegansThe small HSPs in C. elegans consist of proteins ranging from 16-25 kD (Snutch andBaillie 1983; Russnak eta!., 1983). To date, four different genes encoding the 16 kD HSPshave been cloned (Russnak and Candido 1985; Jones et aL, 1986). These are arranged as twodivergently transcribed pairs; one pair, hspl6A, consists of the hspl6-1 and hspl6-48 genes,which are separated by 348 bp of DNA containing the heat shock elements (HSEs) and TATAboxes for both genes (Fig. 1). These genes have been duplicated and inverted at a site 415 bpaway (Russnak and Candido 1985) to produce a large inverted repeat with four genes in 3.8 kb.The hspl6-2 and hspl6-41 genes reside at a second site, hspl6B, and contain their264 I.Hsp 16-(17’ 16-2tATA USE USE ATA’NN..—1—J i i i 1J4 16-41 XbI 16-2A.Chapter I: IntroductionLocus lisp 16—1/46 ( VI Inverted Repent (1) Inverted ret I4.. 4Hsp 16-I 85p16-4 H5p16—48 Hsp 16—ILocus I-Isp 15—2/418.C.I kbFigure 1. Organization of the hspl6 loci of C. elegans. Re-printed from Candido et al. (1989).The open boxes represent introns while the filled boxes represent exons. The direction oftranscription is indicated by the arrows. HSE, Heat shock element.27Chapter 1: Introductionregulatory elements within a 346 bp intergenic region; unlike the hspl6-48/1 genes, theseare present as single copies (Jones etah, 1986). Northern analysis indicates that expressionof the hspl6 genes is absolutely dependent upon heat shock (Russnak et al., 1983). Thepolypeptides encoded by the hspl6-1 and hspl6-2 genes are 145 amino acids in length andshare 92% identity; similarly, those of hspl6-48 and hspl6-41 are 143 residues in lengthand are 94% identical to each other. In contrast, the members of a given gene pair are lessthan 70% identical to each other (Jones et aL, 1986), and the identical residues are allencoded within the second exon; thus each gene pair encodes one of each class of HSP16s, whichis defined by the sequence of exon 1. The HSP16s show a high degree of sequence similarity toboth the small HSPs of Drosophila and to the a-crystallins of the vertebrate eye lens (Ingoliaand Craig 1982; Southgate eta!., 1983).Recently antibodies raised against a peptide corresponding to the carboxyl terminus ofthe HSP16-41 protein (anti-41/125-143) were found to react strongly with a 16 kD bandand weakly with an 18 kD band on Westerns from heat shocked nematodes (Hockertz et al.,1991). Furthermore, antibodies against peptides corresponding to the carboxyl end of theHSP16-2 protein (anti-2/110-145) or to the HSP16-1 protein (anti-1/33-50) detected an18 kD band on Westerns. The anti HSP16-2 antibody also reacted weakly with a 16 kD band.These results suggest that the 16 and 18 kD heat shock proteins of C. elegans are structurallyrelated and raises the possibility that the hspl6-1 and hspl6-2 genes may actually encodethe 18 kD heat shock proteins. Western blot analysis of a two dimensional gel using an antiHSP16-2 antibody as a probe detected 13 heat induced polypeptides, all within the 16-18 kDregion. Thus, the HSP16 and HSP18 polypeptides appear to exist in vivo in multiple isoformswhich presumably arise from post-translational modifications. Purified native HSP16!18proteins formed large aggregates of approximate mass of 400-500 kDa (Hockertz et al.,1991). This suggests that roughly 25-31 monomers of small HSPs potentially form thesecomplexes, comparable to the stoichiometry of 32 monomers inferred for the murine HSP25protein (Behlke et al., 1991).28Chapter 1: IntroductionThe intergenic regions of the hspl6-48/1 and hsplo-41/2 gene pairs are highlyconserved (86% identity) suggesting that elements other than the HSEs and TATA boxes may beimportant for their regulation. In particular, a stretch of alternating pyrimidines and purinesand an inverted repeat may be candidates for control elements (Jones et al., 1986). Kay eta!.(1986) attempted to define the roles of these sequences in hspl6 gene expression bytransfecting mouse fibroblasts with various constructs of the hspl6A locus. Theseexperiments demonstrated that a single HSE could induce bidirectional expression of the hspl6-48/1 gene pair upon heat shock and that the number of HSEs present greatly affects the strengthof the promoter. Disruption of the purine-pyrimidine stretch had no effect on expression (Kayet al., 1986). However, these experiments were performed in a heterologous system and it ispossible that the mouse transcriptional apparatus may not have recognized certain controlelements in the nematode genes, other than the highly conserved HSEs.Although the hspl6A and hspl6B intergenic sequences are 86% identical, the gene lociare differentially regulated (Jones et al., 1989) and display differences in chromatinstructure during heat shock (Dixon et aL, 1990). Nuclear run-on experiments determinedthat the transcription rates of the hspl6 genes are similar. However, while the hspl6-41/2locus continues to transcribe messages at a high level after two hours of heat shock,transcription from the hspl6-48/1 locus begins to decrease within an hour. This results inapproximately a 14-fold difference in mRNA levels between the two loci in favour of hspl6-4 1/2. The disparity in mRNA levels is greatest at the embryo and Li stages, suggesting theexistence of developmental differences in hspl6 gene expression ( Jones eta!., 1989). Thetwo gene clusters also exhibit differences in the timing of appearance and disappearance ofDNAsel hypersensitive sites upstream of the HSEs during heat induction and recovery (Dixon eta?., 1990). These observations suggest there is some differential regulation of the two hspl6gene clusters.29Chapter I: Introduction9. The present studyIn recent years, the volume of information regarding the function and regulation of thehigh molecular weight HSPs has increased dramatically. In contrast, still very little is knownregarding the expression and function of the small HSPs, particularly those which are requiredonly upon heat shock. The strict heat inducibility and the differential regulation observed in thehspl6 gene family of C. elegans inspires a more careful and detailed analysis of these processesin this organism itself. Unfortunately, transformation of gene constructs designed in vitro intothe homologous system was not a technique available to the C. elegans researcher until 1986(Fire , 1986). My arrival in the laboratory in 1987 was fortuitously coincident with theaccessibility of this technique and I consequently set out to evaluate the patterns of hspl6 geneexpression in vivo.30Chapter I: MethodsB: MATERIALS AND METHODS1. Construction of hspl6-IacZ fusionsThe IacZ gene fusions described in this study were constructed by my collaboratorsDonald Jones and Dennis Dixon, and by myself. The vector pPCZ1, abbreviated as 48.1 C in thetext, was constructed by Donald Jones. He inserted a 3500 bp Hindlil-Aflil fragmentencompassing the IacZ gene (nucleotides 18 to 3518 of the expression vector pPD16.43V received from A.Fire, Carnegie Institution) by blunt end ligation into the Hpa I site (nucleotide3565 of the published sequence) of the hspl6-1 gene. Plasmid pPCZ1 contains a completehspl6-48/1 gene pair (hence 48.1C) extending from a Bc! I site at nucleotide 2280 to theV Barn HI site at nucleotide 4186 in the published sequence (Russnak and Candido 1985).Subsequently I microinjected pPCZ1 and established transgenic lines. Construct 41 .2C is theabbreviation for pDX1 6.31, and was the gift of Dennis Dixon who had inititiated expressionstudies of the hspl6-41/2 gene pair while at the Carnegie Institute of Washington. Plasmid41.2C contains IacZ as a 3200 bp XbaI-Stul fragment in the Hpa I site of hspl6-2(nucleotide 1690, Jones et aL,1986). The latter was contained in an Eco RI (nucleotide 540)to Mbo I (nucleotide 2870) fragment encompassing the hspl6-41/2 gene pair. Forcomparative analysis of hsp 16-48/1 with hspl6-41/2 I microinjected 41.2C andestablished transgenic lines. Plasmid pPC16.48-1, referred to as 1.48E1 (48E1 representsfusion to exon 1 of hspl6-48), was constructed by inserting a Sau 3A fragment extending fromnucleotides 987 to 1440 (Russnak and Candido 1985) into the Barn HI site of the nematodeexpression vector pPD16.51 (Fire ot aL, 1990), such that the hspl6-48 promoter wasproximal to IacZ. pPC16.1-48, abbreviated as 48.1E1, contains the Sau 3A fragment in theinverse orientation such that the hspl6-1 promoter is closest to IacZ. Both pPC16.48-1 andpPC16.1-48 were constructed by Don Jones. pPC16.1-48XBAI, abbreviated as 48.1XBE1 inthe text, was constructed by inverting the XbaI fragment (nucleotides 1134 to 1289) betweenthe two HSEs of the hspl6-48/1 locus in the vector pPC16.1-48. A Sau 3A fragmentextending from nucleotides 1121 to 1561 of the hspl6-41/2 locus was cloned into the Barn HI31Chapter I: Methodssite of pPD16.51 such that the hap 16-41 promoter was proximal to IacZ, generatingpHS 16.25 (abbreviated as 2.41 El and also the gift of Dennis Dixon).Plasmid pPC1 6.48-51, abbreviated as 48P, is a transcriptional fusion consisting ofan Mnl I fragment (nucleotides 3085 to 3262) of the hspl6-48/1 intergenic region clonedinto the Hind! site of the pPD16.51 polylinker. Since Mnl I cleaves DNA at a different site toits recognition sequence the resulting construct was analysed by the double stranded dideoxysequencing method of Gaterman et al. (1988) to determine the exact sequence at the hspl6-48/pPD1 6.51 junction. The sequence of the non-coding strand at the junction was lacZpPD16.51 Hincll-GTC/GATTGGCTTATATACCC-hspl6-48. Plasmid pPC16.41-51, referred toas 41 P in the text, is a transcriptional fusion consisting of a Taq I fragment extending fromnucleotides 1169 to 1409 in the hspl6-41/2 intergenic region, inserted at the Acc I site ofpPD16.51; its expected and confirmed junction sequence is lacZ- pPD16.51 AcciGT/CGAAGTTTTTTAGATGCACTAGAACAA-hSp16-41. Subsequently, all of these constructs weremicroinjected into C. elegans oocyte nuclei and transgenic lines established. A complete list ofall strains generated and their genotypes is presented in Appendix.2. Bacterial TransformationsFusion constructs were transformed into competent DH5a cells which had beenpurchased from BRL or prepared locally by the method of Hanahan (1983). Typicaltransformation efficiencies were in the range of iO7iO8 per .tg of plasmid. Transformationmixtures were plated on YT Ampicillin plates and incubated overnight at 37°C for colonygrowth.3. Double stranded sequencing of lacZ fusionsPlasmid DNA was isolated from E. coil for sequencing using a modified version of amethod devised by Gaterman et a/.(1988). Bacteria (1.5 ml) were pelleted at 10,000 rpm(8160 RCF) for one minute, resuspended in 100 jil of lysis buffer ( 8 % sucrose; 5 % Triton32Chapter I. MethodsX-100; 50 mM EDTA ; 0.5 mg/mI lysozyme) and boiled for 2 minutes. Cellular debris waspelleted by centrifugation at 4°C for 15 minutes. After removal of the pellet, the DNA wasprecipitated from the supernatant with 100 p.1 of cold isopropanol followed by anothercentrifugation at 4°C. The DNA pellet was washed, dried and resuspended in 40 jil of dH2O.Typically this procedure yielded 20 jig of DNA.Prior to sequencing, the plasmid was denatured: 8 i.tl of plasmid DNA ( approx. 4 jig) wasmixed with 1 pmol of M13 universal forward or reverse primer, 11 p.1 of dH2O, and 2 p.1 of 2MNaOH , and boiled for two minutes. Upon removal from the boiling bath the denatured DNA wasprecipitated with 2 p.1 of 3M NAQAc, pH 6.8 and 50 p.1 of 95 % ethanol and incubated for 10minutes at -20°C . Subsequently the DNA was pelleted, washed with 70 % ethanol, dried andresuspended in 7 p.1 of sterile dH2O.To anneal the primer to the template 2 p.1 of 5 X sequencing buffer (200 mM Tris, pH7.5; 100 mM MgCl; 250 mM NaCl) and an additional 1 pmol of primer were added, themixture heated at 65°C for 2 minutes and then allowed to cool to room temperature graduallyover the course of about 30 minutes. Sequencing utilized commercial nucleotide mixes (USBiochemical), 35SdATP and modified T7 DNA polymerase (Sequenase version 2.0-USBiochemical) and followed procedures outlined by the manufacturer. Samples were boiledprior to loading on a 6 % polyacrylamide gel and electrophoresed at 20 mA (1500.volts). Sincethe weak 35 f3 emissions cannot penetrate through the urea in denaturing polyacrylamide gels,gels were washed in a solution of 5 % methanol; 5 % acetic acid to remove the urea before beingdried and exposed to film.4. Maintenance of strainsC. elegans strains were maintained on NG plates seeded with E. coil strain 0P50 at20°C, essentially as described by Brenner (1974).33Chapter 1: Methods5. Establishment of transgenic C. elegans strainsDNA was prepared for injection by a modified alkaline lysis procedure (Birnboim andDoly 1979) which included a lithium chloride precipitation step to remove RNA (Sambrook etal., 1989). The DNA was finally suspended in an injection solution consisting of 2 %polyethylene glycol 6000; 20 mM potassium phosphate; 3 mM potassium citrate-pH 7.5 (Fire,1986). Injection needles were pulled from glass capillaries (Cat. No. 1B100F-6, WorldPrecision Instruments Inc., New Haven, Conn.) using a Frederic Haer & Co. micropipettepuller, and filled with equimolar amounts of the selectable and test plasmids (40-200ng/jil ofeach). Wild type C. elegans oocytes were injected at a magnification of 400X using a ZeisslM35 microscope equipped with Nomarski optics, and a Leitz micromanipulator according tothe method of Fire (1986).6. Selection of Transformed ProgenyProgeny transformed with the unc-22 antisense vector, pPD1O.41 were identified byvisual inspection for twitching (Moerman and Baillie 1979). Strong unc-22 mutants shakeconsiderably but to retrieve weaker phenotypes transgenics were selected. in the presence of 1% nicotine (Moerman and Baillie, 1979). Selection of progeny transformed with pRF4, aplasmid containing the rol-6 dominant allele sulOOG, was simply by visual inspection(Kramer et al., 1990; Mello, personal communication) . The antimorphic allele sulOO6encodes a mutant collagen and these mutants possess an altered body cuticle which forces theanimal to roll onto its right side.Each selected transgenic animal was placed on a separate plate for propagation. A linewas considered established after the selectable phenotype had been sucessfully propagated forthree generations.34Chapter 1: Methods7. Viable freezing of transgenic strainsNematodes were preserved for long term storage by viable freezing in glycerol.Generally, animals were washed off a 9 x 1.5 cm plate almost devoid of food with cold 0.14 NaCIand centrifuged briefly at low speed (about 3000 rpm). The worm slurry was suspended inapproximately 1 ml of M9 buffer and an equivalent volume of sterile Basal S containing 30 %glycerol was added. Aliquots of 400- 500 il were frozen overnight in a -70 freezer beforebeing transferred to liquid nitrogen for long term storage. Frozen stocks were thawed at roomtemperature for approximately 10 minutes before they were spotted on NG ; 0P50 plates.Typically, survival rates of larval stages ranged between 60 % and 80 % while, survivingadults were rendered sterile by the procedure and embryos failed to survive at all.8. Heat Shock ConditionsInitially, the temperature response of selected transgenic lines was investigated todetermine a suitable standard temperature for heat shock experiments. Transgenic linesmaintained at 20°C were subjected to heat shocks of 25°C, 27°C, 29°C, 31°C and 33°C for twohours and then allowed to recover at room temperature for 15 m.inutes before staining for f3-galactosidase activity. These experiments established that 33°C consistently induced theexpression of transgenes. Routinely, transgenic lines were tested for IacZ expression .by heatshocking at 33°C for two hours on pre-warmed .NG plates spread with bacteria. A double heatshock consisted of two two-hour exposures at 33°C separated by a 30 minute recovery period at20°C. Worms were subsequently allowed to recover for 15 minutes at 20°C before beingwashed off in distilled water, permeabilized by lyophilization and acetone treatment, andincubated in a histochemical stain containing 0.2 M NaPO4-pH7.5, 1 mM MgCl, 5 mMpotassium ferricyanide, 5 mM potassium ferrocyanide, 0.004 % SDS, 0.075 jig/jil kanamycinsulphate, 0.002 .tg/il DAPI, 0.24 g/3Il X-gal (Fire et at., 1990) overnight at roomtemperature. Staining was often visible after only 15 minutes of incubation. As a control,- some worms of each strain were incubated in staining solution without prior heat shock35Chapter I: Methodstreatment. No staining was ever observed in strains maintained at 20°C . Only three of 71transgenic lines did not stain after heat shock, and in one of these instances Southern analysisindicated that the lacZ fusion had not been co-transformed with the rol-6 sulOO6 plasmid(data not shown). After initial characterization, certain strains were selected for furtheranalysis and subjected to heat shocks of varying duration. In addition, to study developmentalstages, strains were synchronised by bleaching gravid adults to obtain eggs (Emmons et aL,1979).9. Identification of f3-galactosidase staining cellsTransgenic animals were washed in distilled water, permeabilized by lyophilization andacetone treatment and incubated in a stain containing 5-bromo 4-chloro 3-indolylgalactosidase(Xgai) as described by Fire et al. (1990). Animals were permanently mounted in 80 %glycerol, 20 mM Tris-pH 8.0, 200 mM sodium azide prior to microscopic examination.Nuclei were scored for the presence or absence of blue stain. The identity of stained celltypes was determined by the size and shape of the nucleus and its position in the animal relativeto other nuclei, as defined by Sulston (1976), Sulston and Horvitz (1977), Albertson andThomson (1976), and Sulston et a!. (1983). Counterstaining with DAPI (4,6-diamidino-2-phenylindole) allowed visualization of all nuclei in an animal using fluorescence microscopy(Ellis. and Horvitz 1986). Frequently the staining of j3-galactosidase activity was so strongthat complete quenching of the DAPI was observed. This proved to be an asset as it allowedidentification of some nuclei by exclusion. When cell identity could not be confidentlydetermined, the staining nucleus was not scored.10. Preparation of Transgenic Genomic DNAHigh molecular weight transgenic DNA was isolated using a modified procedure ofReymond (1987). This method allowed for the rapid purification of DNA from numeroustransgenic strains simultaneously. Nematodes were washed off crowded plates with cold 0.1436Chapter 1: MethodsNaCI and collected in a microfuge tube. Usually a worm slurry of about 100 jil was used perDNA preparation. Digestion solution (200 il of 50 mM Tris, pH 7.5; 20 mM EDTA; 1 mg/ mlProteinase K) was added to the nematode slurry, mixed and then incubated at 65°C for 30 to 60minutes. Subsequently, 200 jil of formamide was added to the microfuge tube and the mixturewas incubated on ice for 10 to 15 minutes. Addition of 100 jil of 8M NH4OAc was followed byanother 10 minute incubation on ice. After RNA and cellular debris were pelleted bycentrifugation at 10,000 rpm for 3 minutes, the supernatant was precipitated with 1 ml of cold95 % ethanol and incubated on dry ice for 10 minutes. The DNA was pelleted by centrifugationat 12,000 rpm (11,750 g) for 5 minutes and the pellet resuspended in 400 jLl of TNE (5 mMTris, pH 7.5; 5 mM NaCI; 0.2 mM EDTA). Addition of 50 jtl of 8M NH4OAC and 1 ml of 95 %ethanol re-precipitated the DNA, and after incubation on dry ice and centrifugation, the pelletwas finally resuspended in 50-100 p.1 of TNE.DNA samples were quantified spectrophotometrically by absorbance at 260 nanometers(Sambrook et al., 1989). Typically, 50 to 100 jig of DNA was isolated from one plate (9 cm x1 .5 cm) of worms. The quality of the DNA samples was analysed by electrophoresis in 1 %agarose gels.11. Southern Transfer and Analysis of Transgenic DNAsDigested DNA was transferred to nylon membranes, Zetaprobe (Bio-Rad) or Hybond-N(Amersham) by Southern blotting (Southern, 1975) modified according to the manufacturers’instructions. After transfer was complete, membranes were rinsed briefly in 2 X SSC and driedat room temperature. Zetaprobe membranes were used without further treatment whereas DNAwas permanently fixed to Hybond-N membranes by U.V. crosslinking.12. DNA Dot Blot ProceduresGenomic and plasmid DNA was transferred to Hybond-N membranes using a DOT blotapparatus (Bio-Rad) and applied vacuum according to the following procedure. Six hundred ng37Chapter 1: Methodsof DNA was suspended in 50 il of TE (10 mM Tris, pH 8.0; 1 mM EDTA) and denatured by theaddition of 0.1 volumes of 3M NaOH. The mixture was incubated for one hour at approximately65°C and then cooled to room temperature. Subsequently, one volume of 2M NH4 QAC, pH 7.0was added. Prior to application, the samples were diluted to 400 p.1 with 2M NH4OAc. Allsubsequent serial dilutions used 2M NH4OAc. The membrane was prepared for installation inthe apparatus by hydration in deionized water followed by a rinse in 1 M NH4OAc, pH 7.0. Priorto sample application the wells were rinsed with 500 p.1 of 1M NH4OAc, pH 7.0. After transferwas complete the membrane was rinsed briefly in 2 X SSC and dried at room temperature.Hybridization procedures followed those previously described. Signal quantitation was bydensitometry on a LKB Ultroscan XL laser densitometer. Copy number in transgenic strains wasdetermined from points which fell on the linear part of the standard curve derived fromdilutions of plasmid DNA.13. Labelling of Radioactive ProbesProbes of high specific activity were prepared for hybridizations by either nicktranslation or primer extension. Nick translations of pPD16.43 followed protocols described inSambrook et al.(1989). Typically probes were labelled with 32PdATP and 32PdCTP to aspecific activity between 0.5 and 1 x108 cpmlp.g using this method. Primer extensionlabelling of M13 templates containing the entire IacZ gene followed procedures outlined byRussnak and Candido (1985). This method generated probes with specific activities ofapproximately 108 cpm per jig of template. DNA molecular weight markers were created byend labelling lambda DNA that had been digested with Hind Ill. In this procedure 1 jig oftemplate was incubated at room temperature for 10 minutes in a solution of 10 mM Tris, pH7.5; 10 mM MgCl; 1 mM DTT; 10 jiCi of 32PdATP and 5 units of E. coil DNA polymeraseKlenow fragment. Addition of EDTA to a final concentration of 20 mM stopped the reaction.All radiolabelled probes were purified by chromatography on G-50 Sephadex spincolumns. Usually 10,000 cpm of radioactive molecular weight marker DNA was loaded directly38Chapter I. Methodsinto slots of agarose gels prior to electrophoresis. For most experiments, cpm of nicktranslated or primer extended probe was added per ml of hybridization solution.14. Hybridization ConditionsUsually membranes were incubated in 10 to 15 ml of hybridization solution for 4 hoursprior to addition of the radioactive probe. Hybridization solution for Zetaprobe membranesconsisted of 1 % SDS; 0.5 % BLOTTO; 50 % formamide; 4 X SSPE. Hybridization solution forHYBOND membranes consisted of 5 X SSPE; 5 X Denhardt’s; 50 % formamide; 0.5 % SDS. Insome experiments 20 il of 10 mg/mI sheared single stranded calf thymus DNA was added to thehybridization to reduce non-specific binding.After overnight incubation at 42°C membranes were washed in SSC and SDS with thefinal wash consisting of 0.1 x Ssci 0.1 % SDS at 50 °C. Subsequently membranes were exposedto X-OMAT-AR film (Kodak) for autoradiography.39Chapter 1: ResultsC: RESULTS1. Construction of hspl6-lacZ fusions and selection of transformantsThe hspl6-lacZ fusions used for transformation are described in Fig.2. For ease ofreference, the plasmids will be referred to using an abbreviated terminology. Full designationsfor the constructs are given in “Methods”. Constructs 48.1C and 41.2C contain the completehspl6-48/1 and hspl6-41/2 gene pairs, and are in-frame translational fusions of the IacZcoding region with the second exon of hspl6-1 and hspl6-2, respectively. Constructs 48.1 El,1 .48E1 and 2.41 El are translational fusions to lacZ in which a Sau3A fragment containing theintergenic region of a gene pair was fused in-frame to lacZ at the Sau3A site in exori 1 ofhspl6-1, hspl6-48 and hspl6-41, respectively. Plasmid 48.1XBE1 is a translationalfusion in which the Xba I fragment of the 48.1 El promoters was inverted. Finally, 48P and41 P are transcriptional fusions containing the promoter (i.e. TATA box, HSEs and someupstream sequence) of hspl6-48 and hspl6-41, respectively. All constructs contained theSV4O nuclear localization signal fused to the beginning of the lacZ coding region (Fire 1986).Transgenic C. elegans strains were constructed by microinjecting oocytes of wild-typehermaphrodites with DNA of the desired construct, together with a selectable marker.Selection of transformants co-injected with pPD1O.41, an unc-22 antisense vector (Fire etal., 1991), was by visUal inspection, as unc-22 mutants display a twitching phenotype;alternatively, animals were examined in the presence of nicotine, which enhances twitching andallows selection of weaker phenotypes (Moerman and Baillie 1979). Animals which had beentransformed with rol-6 sulOO6 were identified by their right rolling phenotype (Kramer etal., 1990; MeIlo, personal communication).The transformed lines thus obtained transmitted the marker phenotype at a frequency of20-95%.40Chapter I. ResultsFigure 2. Construction of hspl6-IacZ transgenes. 48.1C consists of a complete hspl6-48/1gene pair, including the 5’ and 3’ non-coding sequences of both genes, with the E. coil IacZgene inserted in-frame into a unique Hpal site in the second exon of hspl6-1. 41.2C is thehomologous construct using the hspl6-41/2 gene pair. The exon 1 fusions l.48E1, 48.1 Eland 2.41 El were constructed by cloning a Sau3A fragment containing the intergenic sequence ofhspl6-48/1, hspl6-1/48 and hspl6-2/41, respectively, into the BamHl site of the IacZexpression vector pPD16.51 (Fire et at., 1991, in press). These constructs include the first15 (2.41E1 and l.48E1) or 17 (48.1E1) amino acid residues of the respective HSP16. In48.1XBE1 the Xbal fragment between the HSEs of 48.1E1 was inverted. The polyadenylationsignal following the IacZ gene in plasmid pPD16.51 is derived from the 3’ non-coding region ofthe myosin gene, unc-54 (Epstein 1974; Fire 1fiLL, 1990). The arrows indicate thedirection of IacZ transcription. 48P and 41P are transcriptional fusions which remove theHSEs and TATA boxes of the hspl6-1 and hspl6-2 genes, respectively, but retain a singlepromoter (hspl6-48 or hspl6-41). NLS, SV4O nuclear localization signal.41Chapter 1: ResultsTranslational Fusions (Complete):I____ _____I IN LSI LACZTATA HSE Z HSE TATAHspl6-48 48.1C Hspl6-1Hspl6-41 41.2C Hspl6-2Translational Fusions (Exon 1):Hspl6-48 48.1E1 Hspl6.1Hspl6-1 1.48E1 Hspl6-48Hspl6-2 2.41E1 Hspl6-41.Hspl 6-48/Xba-Hspl 6-1—48.1 XBE1—Hspl 6-48-Xba/Hspl 6-1Transcriptional Fusions:Z HSE TATAHspl6-48 48PHspl6-41 -41P42Chapter 1: Results2. Expression of the hspl6-IacZ transgene is temperature dependent.Transgenic lines were heat shocked for two hours and stained with a solution containingXgal (see Methods) to assay for expression of the hspl6-IacZ fusion genes. Figure 3 showsstaining of a transgenic line carrying the 48.1C transgene before and after heat shock. Nostaining was observed in control worms, indicating that expression of the transgene wasdependent upon heat shock. Eleven lines carrying the 48.1 C transgene were examined and allgave identical results. Not all heat shocked worms within a strain stained, presumably due toincomplete germ line transmission of the extrachromosomal array. Moreover, mosaicism wasfrequently observed among the animals which did stain. This most likely results from randomloss of the array during mitosis.To determine if the selection plasmid phenotype absolutely correlated with incidence ofIacZ expression, rollers or twitchers and phenotypically wild type animals from the same linewere picked to separate plates, heat shocked for two hours and then stained with Xgal. In allthree strains tested, 100 % of the selected rollers and twitchers stained after heat shock. Intwo of the three strains (one roller and one twitcher), one third of the phenotypically wild typeanimals also stained after heat shock, even though the twitcher strain had been screened in thepresence of nicotine, implying that f-galactosidase activity is a more penetrant phenotype.Since the 3-gaIactosidase assay detects enzyme activity by a sensitive histochemical procedure,it is conceivable that a greater quantity of therol-6 sulOO6 gene product may be required toproduce abnormal cuticle and that more unc-22 antisense RNA may be needed to interfere withthe wild type product to mimic loss-of-function alleles. Alternatively, since each of the wildtype selection genes are expressed in specific tissues, the extrachromosomal array may havebeen lost in those particular tissues during development but not in others.43Chapter 1: ResultsFigure 3. Heat shock dependence of IacZ expression. In ifli staining of -galactosidase activityin PC6, a strain carrying an extra-chromosomal genetic element composed of mixed arrays of48.1 C and the rol-6 selection vector pRF4, is shown. Magnification 100 X. a) Staining of amixed population of animals without prior heat treatment. b)Staining following a two hour heattreatment at 33°C. Staining of intestinal, body muscle, hypodermal and pharyngeal nuclei isclearly visible.444Lfl0.1’•e-—Ir/J-—r,‘1-’qi-‘.4,.-oChapter I: Results3. Establishing Standard Heat Shock Conditions for ExperimentsTemperature studies of transgenic lines established that the minimum temperature atwhich any staining was observed was typically between 29° and 31°C, but staining was sporadicand infrequent over this temperature range (Figure 4). In one strain (PC6), 1% of embryosstained at 25°C. Also, on one occasion an adult stained moderately at 25°C (Figure 4). By 31°Can extensive expression pattern was observed in PC6 animals. However, somatic expressionin PC2O animals remained patchy at this temperature (Figure 4). Hence, 33°C was chosen as asuitable heat shock temperature since it provided good expression of the IacZ transgene in allstrains.4. The transgenes 48.1C and 41.2C are expressed in a tissue general mannerExpression of both the 48.1C and 41.2C transgenes was consistently observed inembryos from gastrulation onward, in all larval stages, and in adults. In post-embryonicstages, both transgenes were expressed in most tissues (Fig. 5). Curiously, the nuclearlocalization signal (NLS) in the 41.20 element did not effectively target -galactosidase to thenucleus, even though the IacZ gene had been inserted into hspl6-2 of 41.20 at the analogousposition (Fig. 2)10 the insertion in the hspl6-1 gene of 48.1C. This suggests that the NLS of41 .2C is non-functional in the -galactosidase fusion protein, while remaining active in the48.1C fusion protein.While the 41 .2C transgene is expressed in numerous tissues especially the pharynx andgut, it is not expressed in the gonad. 48.1C transgenic animals also express the transgene in atissue general fashion, and in this case the nuclear localization of the stain greatly facilitatescell identification. Examples of every cell type have been observed in strains carrying thistransgene, including neurons, muscle, intestine, and hypodermis. Only the germ line failed toexpress this fusion gene. Figure 6 shows some examples of the tissue distribution of expressionseen with these constructs. Tissues expressing the hspl6-lacZ fusion in these strains includenerve ganglia in the head and ventral cord; hypodermal nuclei of the lateral hypodermis, vulva,46Chapter I: ResultsFigure 4. Expression of -gaIactosidase is proportional to temperature in hspl6-IacZtransgenic animals. Cultures of PC6, a 48.1C/pRF4 strain and PC2O, a 2.41E1/pPD1O.41strain were heat shocked on pre-warmed plates at varying temperatures (25°C, 27°C, 29°C,31°C, and 33°C) for two hours, recovered for 15 mm. at room temperature, and then assayedfor f3-galactosidase activity.47I0I•%4•;co.4-I000p-I.I..oo:•“/A,C.0DI’4-00UF0(-‘a%.01C’4C)49•100027°C0.000025°C29°CPC2 000I I• .‘7031°C/‘74I‘31CChapter 1: ResultsFigure 5. LacZ expression of the 48.1C and 41.2C transgenes in response to heat is tissuegeneral. a) Brightfield image at 200X magnification of an adult hermaphrodite carrying the48.1C transgene. Blue precipitate is seen in body muscle, intestinal, hypodermal, neural andpharyngeal nuclei. b) Staining in a late gastrulation/ comma stage embyo. Thirty-six stainingnuclei were counted. Magnification 1 250X. C) Unlocalized intracellular expression in a 41 .2Cadult hermaphrodite. Magnification 200X. Although the stain is diffuse, expression is evidentin the intestine, pharynx, vulva, and ventral hypodermis. The germ cells of the gonad areclearly not stained.50S.:S.a•14’-‘?;‘14I*•‘aS.•f0bChapter I. ResultsFigure 6. Cell types expressing hspl6-IacZ transgenes upon heat shock, a) Staining of nucleiof the lateral hypodermis (hyp) in a 1 .48E1 animal. b) Body muscle (bm) nuclear expressionin a strain carrying the 48.1 C transgene. Stained intestinal and hypodermal nuclei are visiblein the background, out of the plane of focus. C) Expression in the head of a 2.41E1 animal.Neurons of the nerve ring (nr) are stained, as well as those of the ventral and retro-vesicularganglia. The pharyngeal-intestinal valve (piv) as well as muscle, epithelial and marginalnuclei in the terminal bulb (tb) are indicated. d) Expression of the 48.1C transgene in anadult male. In addition to general body muscle and hypodermis, V and T derived nuclei of therays are stained. e) Expression of the 48.1C transgene in neuronal nuclei of the ventral cord(vc). f) A coelomocyte (cc) in the pseudocoelomic cavity of a 48.1 C adult hermaphrodite. g)Vulval expression, lateral view. P derived hypodermal nuclei of the vulva as well asmesodermal tissue above the vulva (vu) are stained. h) Vulval expression, ventral aspect.Neuronal and hypodermal nuclei of the P lineage (p1) are indicated.52aChapter I. Resultshead and tail; body wall and pharyngeal muscle, somatic gonad, mesodermal tissue such ascoelomocytes, and muscles involved in defecation (the latter not shown). These results indicatethat the hspl6-48/1 and hspl6-41/2 gene pairs are both expressed in a heat inducible,tissue general manner in vivo and that individual cells are capable of expressing both genepairs. On the other hand, subtle differences in the patterns of expression between the twoconstructs were observed; in particular, 41.20 animals tended to stain most intenselythroughout the pharynx while 48.1 C animals demonstrated conspicuous expression in body wallmuscle. The latter characteristic was never unequivocally observed in 41 .2C animals.However, these differences were difficult to document since 41 .2C animals displayed diffusestaining. Thus I examined a different series of IacZ fusions of these gene pairs in order toassess potential differences in expression more accurately (see below).5. Quantitative differences in the tissue specific expression of the hspl6 gene pairs.In order to determine if differences exist in the tissue distribution of the hspl6-48/1and hsp 16-41/2 gene products, I established 8 transgenic lines containing the fusion 1.48E1and 8 containing 2.41E1 (Fig. 2). In addition, to assess whether the orientation of theintergenic region relative to IacZ affected expression, I also established 6 lines carrying48.1 El, the inverse of 1.48E1. Initially, all of these strains were heat shocked for two hoursat 33°C, stained and qualitatively analyzed. The typical patterns of expression observed in theselines are shown in Figure 7. Again, each transgene was expressed in a wide variety of tissues.However, certain differences were apparent. For example, all of the strains containing 2.41 Eltended to stain very well in the intestine and pharynx while 1 .48E1 strains consistently showedbetter expression in muscle. Such differences were not exclusive between constructs butrather quantitative in nature. Some 1 .48E1 animals did stain in pharyngeal muscle, but at afrequency lower than that of 2.41E1 animals.To quantify these differences, I selected two or more strains of each construct andsubjected mixed age populations of animals to varying durations of heat shock at 33°C. The54Chapter I: ResultsFigure 7. Expression of hspl6-exonl fusions. Magnification 170X. a) L4 animal carryingthe 1 .48E1 transgene. Staining is intense in hypodermal nuclei of the main body syncitium,head and tail, and in body muscle. Some intestinal expression is also visible. b) Adulthermaphrodite carrying the 48.1 El transgene. Expression is prominent in embryos and in theintestine while hypodermal expression is weak. C) L4 animal carrying the 2.41 El transgene.Staining is intense in the intestine and pharynx, and in nerve ganglia in the head.55LfIw\VChapter 1: Resultsresults of this experiment are summarized in Table 1. Only the tissues showing the moststriking differences are included for comparison. For example, all transgenic lines showedsome expression in the mesodermally derived tissue of the somatic gonad, and in hypodermalnuclei of the vulva, but at frequencies which were not significantly different, so data from thesetissues were not included in Table 1.A number of general observations can be made with regard to the results in Table 1.Firstly, relatively short periods of heat shock were required to induce visible expression of thetransgenes. Even a fifteen minute heat shock without a recovery period was often adequate togive visible expression of 1-galactosidase in some strains. Longer periods of heat shock,however, increased the proportion of tissues staining. Mosaicism of expression was stillobserved, however, even after a double heat shock, since 100% staining of all tissues was neverobserved. It should be noted that tissue expression was simply scored for the presence orabsence of any staining, i.e. all nuclei of a given tissue were not necessarily stained.While all constructs showed a high frequency of intestinal staining at longer heat shocks,the frequency of expression dropped to 67% in 1 .48E1 animals versus 83% and 95% for48.1E1 and 2.41E1 animals, respectively, for 15 minute heat shocks. A striking differencewas observed in pharyngeal expression. Both 48.1E1 and 1.48E1 transgenics stainedinfrequently in the pharyngeal intestinal valve. 2.41 El worms, in contrast, consistentlyexpressed the transgene in this tissue as well as in the rest of the pharynx, includingpharyngeal muscle, marginal and epithelial cells. Expression in the pharynx for the otherconstructs was less consistent and often limited to the gland nuclei of the terminal bulb. Neuralexpression was most prevalent in the 2.41 El transformants and often included ganglia of thenerve ring, the ventral and retrovesicular ganglia, and the lumbar and preanal ganglia in thetail. Neural expression in 48.1 El animals was usually limited to head ganglia and was lessprominent than in 2.41 El transformants; it was rarely observed in 1 .48E1 worms. However,expression in the latter animals was difficult to document due to the large number of intenselystaining hypodermal and body muscle nuclei in the head.57Chapter 1: ResultsTable 1. Tissue distribution of /3-galactosidase staining in transgenic C. elegans heat shockedfor 15-120’ at 33°C.TREATMENT CCNS1TUCT INTEST PIV P1-IA NERVE HYP BM120’ HS 2.41 El 97 90 90 53 70 4030’REC l.48E1 80 20 47 - 77 97120’ HS 48.1 El 90 0 30 30 70 7548.1XBE1 72 0 5 0 45 80120’ HS 2.41 El 93 77 83 33 50 571.48E1 87 30 80 - 77 9348.1E1 95 0 25 20 60 7048.1XBE1 95 0 50 35 55 8090’HS 2.41E1 90 90 87 40 43 431.48E1 80 20 83 - 73 9048.1E1 95 0 20 10 35 5548.1XBE1 100 15 75 30 35 7560’ HS 2.41 El 93 70 87 30 53 531.48E1 73 17 47- 50 8348.1E1 95 20 55 40 45 6048.1XBE1 100 20 85 25. 40 7530’ HS 2.41 El 100 20 53 7 27 201.48E1 67 7 33 - 57 9048.1E1 100 0 35 25 25 7048.1XBE1 95 0 40 20 40 7515’HS 2.41E1 95 13 23 3 7 0l.48E1 67 0 23 - 43 6748.1E1 83 0 0 0 20 3348.1XBE1 95 0 55 35 25 40Three strains of each of 2.41E1 (P019, 20, and 30) and 1.48E1 (P016, 31, and 33), and twostrains of each of 48.1 El (PC52 and 55) and 48.1 XBE1 (P062 and 67) were selected for thisanalysis. Ten or more animals of each strain were scored for each datum. Numbers representthe percentage of staining transgenic animals which stain in the specified tissue. HS, Heatshock; REC, Recovery; INTEST, Intestine; Ply, Pharyngeal intestinal valve; PHA, Pharynx;HYP, Hypodermis; BM, Body muscle; -, Unable to determine.58Chapter 1: ResultsAll constructs seemed to be expressed at high frequency in hypodermal nuclei followingextensive heat shocks (double or two hour shocks), but expression was far more consistent atshorter heat shock times for the 1 .48E1 transformants. These strains were also far morelikely to show expression in hypodermal nuclei of the lateral bands and of the head and tail.More significant is the prominent body muscle expression consistently observed in 1 .48E1animals as opposed to 2.41E1 and 48.1E1 animals. Thus, while the 2.41 El construct seemed toconfer greater expression in pharyngeal muscle, l.48E1 was expressed much better in bodymuscle.Collectively, these results suggest that there are differences in the priority of tissueexpression between gene pairs. Surprisingly, the inverse construct 48.1 El showedcharacteristics more reminiscent of those of 2.41 El, particularly with regard to intestinal andneural expression. On the other hand, the 48.1 El construct did not express extensively in thepharynx as did the 2.41 El fusion. Thus qualitative as well as quantitative differences inexpression can be seen both between and within gene pairs.I was curious about sequences which might confer tissue specificity for expression.The region of alternating purines and pyrimidines and the potential cruciform structure in thecentre of the intergenic regions were most intriguing. To this end the Xbal fragment of thehspl6-48/1 intergenic region was inverted in the 48.lEl to generate 48.1XBE1 (Fig. 2) inorder to see if the pattern of expression could be altered to more closely resemble that of the1.48E1 construct (Table 1). Generally the 48.1XBE1 animals resembled most closely the48.1 El strains suggesting that the orientation of the sequences between the HSEs has no effecton tissue preference. However, the frequency of pharyngeal expression was slightly elevated in48.1 XBE1 animals, approaching the values observed for 1 .48E1 animals, suggesting that theremay be moderate influences within this region which affect pharyngeal expression.59Chapter 1: Results6. Elimination of one promoter from an hspl6 gene pair significantly reduces somatic tissueexpression without affecting embryonic expression.Multiple HSEs can function cooperatively to induce high levels of heat inducedtranscription (Shuey and Parker 1986; Xiao et a!., 1991). I therefore wished to see ifremoving part of the intergenic region containing the HSEs and TATA motif of one gene wouldaffect the overall expression from the remaining promoter. Using stably transformed mousecell lines, Kay et aL (1986) demonstrated that the level of transcription of the hspl6-48/1gene pair was generally proportional to the number of heat shock elements present in theintergenic region. Since I observed differences in the priority of tissue expression which weredependent on the orientation of the gene pair relative to the reporter gene, it was of interest todetermine if these differences could be attributed to a specific element in the intergenic region.Transcriptional fusions which removed the hspl6-1 (48P) or hsplô-2 (41P) HSEs andTATA boxes from the intergenic regions of the hspl6-48/1 and hspl6-41/2 loci,respectively, were constructed (Fig. 8) and tested.All transformed lines carrying the constructs 48P (4 lines analyzed) or 41 P (4 linesanalyzed) still expressed f3-galactosidase at high levels in embryos (gastrula stage and later),while expression in the post embryonic stages was extremely limited (Fig. 8 and Table 2).Staining in larval and adult tissues was only consistent when long heat shocks were used (twohours or double two hour heat shocks separated by a recovery period). When somaticexpression was observed in 48P transformants, it was often limited to body muscle andhypodermis (64 % of staining worms treated with a double heat shock) although some intestinalexpression was also observed (14 % of animals treated with a double heat shock).Transformants carrying 41 P showed more somatic expression than 48P transformants,but considerably less than 2.41 El strains. Again, much longer periods of heat shock wererequired to give consistently detectable somatic expression, while embryos required only a halfhour heat shock. While the intensity of expression was reduced in 41P animals, thedistribution of tissues affected resembled that of 2.41 El worms. For example, after a double60Chapter 1: ResultsFigure 8. Expression of the 48P transcriptional fusion in heat shocked transgenic animals.Magnification 200X. a) Embryos within the uterus intensely stained even after a short heatshock (33°C, 30 mm.). b) The maximum somatic tissue expression seen in 48P animals.After a double two hour heat shock with an intervening 30 minute recovery period, this animalstained in the lateral hypodermal nuclei. C) Typical 48P hermaphrodite showing minimalsomatic tissue expression after a double heat shock. Staining is seen in two intestinal nuclei, ingland nuclei in the terminal bulb, and in embryos.61(N7‘-àpILChapter 1: ResultsTable 2. Distribution of /3-galactosidase activity in transgenic strains carrying transcriptionalfusions.For each datum, ten animals were scored for staining in the specified tissue unless otherwiseindicated. TRT, Treatment; HS, Heat shock; REC, Recovery; INTEST.,lntestine; HYP.,Hypodermis; PHA., Pharynx; PE, Post embryonic; -, not applicable.TRT. FUSION STRAIN INTEST. NERVE MJSCI..E HYP. PHA. % PE %EGGSPC # SrAc3ES120’HS 41P PC42 14/15 5/15 5/15 9/15 3/15 6% 13%30’REC PC45 14/20 4/20 5/20 6/20 6/20 36 % 40 %120’HS AVE 80 % 26 % 29 % 43 % 34 %48P PC38 4/17 0/17 12/17 2/17 0/17 14% 20%PC4O 0/11 0/11 6/11 3/11 0/11 12 % 36 %AVE 14% - 64% 18%120’HS 41P PC42 7 1 2 9 0 3% 17%PC45 2/1 1 1 7 8 2 19 % 22 %AVE 43% 10% 43% 80% 10%48P PC38 6 0 8 0 0 14 % 35 %PC4O 0 0 0 0 0 0% 21%AVE - - - - -90’HS 41P PC42 7 0 1 5 0 8% 25 %PC45 5 1 7 6 5 8% 4%AVE 60% 5% 40% 55% 25%48P PC38 0 0 1 0 0 11 % 13%PC4O 1 0 1 0 0 <1% 29%AVE -- 1% - -60’HS 41 P PC42 9 3 4 2 1 3 % 7%PC45 3 3 9 0 3 5% 7%AVE 60% 30% 65% 10% 20%48P PC38 1 0 5 0 7 15 % 36 %PC4O 0 0 0 0 0 0% 12%AVE - - - - -30’HS 41P PC42 10 2 4 3 0 6% 8%PC45 0 0 4 6 3 5% 7%AVE 50% 10% 40% 45% 15%48P PC38 0 0 9 2 1 10 % 25 %PC4O 0 0 0 0 0 0% 10%AVE - - - - -63Chapter 1: Resultsheat shock, 80% of 41 P animals stained in the intestine, 29% in body muscle and 43% inhypodermis, values somewhat lower but comparable in distribution to the corresponding 97%,40%, and 70% observed after a double heat shock in 2.41 El worms. Thus the removal of theupstream hspl6-2 promoter sequences and/or the removal of the hsplô-41 translational startsite in the 41P construct resulted in a reduction of the overall intensity and frequency ofexpression but did not seem to restrict tissue distribution. It should be noted that the fragmentof the intergenic region in the 41 P transcriptional fusion is somewhat larger than that of 48P,and that the latter lacks the transcriptional start site of the original gene; this probablyexplains the much better expression of the 41 P construct.7. Southern Analysis of Transgenic StrainsSouthern analysis of genomic DNAs digested with Apa I and Sma I verified the presence ofthe lacZ fusion in transgenic “twitching” strains while no significant hybridization signal wasobserved in control (N2) DNA (Figure 9). However, nick translation resulted in probes whichdid not consistently label to high specificity. For this reason, and to avoid radiolabelling nonlacZ pPD1 6.43 plasmid sequences which might cross hybridize with homologous sequences ofthe selection plasmid (e.g. such as the ampicillin gene sequences), M13 probes containing lacZwere generated by primer extension and routinely used to analyse Southern blots (Figure 10).The presence of numerous hybridizing fragments of various sizes which are smaller than the1.9 kb lacZ band expected by digestion with EcoRl and Ec0RV, coupled with incompletetransmission of the transformed phenotype suggests that the injected DNA formed complexextrachromosomal arrays as described by Stinchcomb et al (1985).While equivalent amounts of DNA were loaded into each slot, the copy number of the lacZtransgene varied widely between strains making copy number estimates difficult. In additionmany DNA samples did not appear to be completely digested, resulting in intense, but nonspecific smearing of signals. Apparently, while the formamide DNA isolation procedureemployed is rapid and produces high molecular weight DNA, inconsistent restriction suggests64Chapter I. ResultsFigure 9. Southern analysis of Smal/Apal digested genomic transgenic strain DNAs probed withnick translated pPD16.43. The solid arrow indicates the IacZ containing fragment; the cleararrow, the fragment containing other vector sequences. The latter fragment is larger intransgenic strains due to the included hspl6 sequences. Lanes: M, Lambda DNA digested withHindlll; 1, N2 control; 2, PC1A, a 48P/pPD1O.41 strain; 3, PC3A, a 48P/pPD1O.41 strain;4, PC9A, a 48P/pPD1O.41 strain; 5, PC13A, a 48P/pPD1O.41 strain; 6, N2 control; 7, 60 pgof pPD16.43; 8, 600 pg of pPD16.43; M, Molecular weight marker. These strains did notexpress the 48P transgene due to lack of a polyadenylation signal in the 3’ end of the originalplasmid vector. Hence this Southern was originally done to check for the presence of the iacZtransgene. These strains were discarded and not included in the subsequent study.65Chapter 1: ResultsMl 23456 789M66Chapter 1: ResultsFigure 10. Southern analysis of transgenic genomic DNA probed with primer extended IacZproducts. Genomic DNAs (3 rig) were digested with EcoRl and Ec0RV, separated byelectrophoresis on a 1 % agarose gel and tranferred to a nylon membrane by the method ofSouthern (1975). The arrow indicates the 1.9 kb band expected by liberation of the IacZfragment by EcoRl/ Ec0RV digestion. Lanes: M, Lambda! HindIll molecular weight marker;transgenic strains as numbered; 1, 10, 100 copies of IacZ contained in pPD16.43.67Chapter I: ResultspPD16.43TRANSGENIC STRAIN: PC NO.M 6 501631 33525562_6738(165 1920a 145 1 1010068Chapter 1: Resultsthat protein contamination is an inherent problem with this procedure. To circumvent theseproblems, dot blots were employed to provide more accurate estimates of transgene copynumber.8. Determination of IacZ transgene copy numberBy DNA dot blot analysis, I have estimated the copy number of the IacZ transgene inthese arrays to be in the range of 5-750 copies per genome equivalent (Fig. 11; Table 3). Thenumber of animals carrying the extrachromosomal array was inferred by selection of themarker phenotype prior to harvesting. Since in some strains one third of the phenotypicallywild type animals may actually possess the array, the copy number results are probablyoverestimates.In practice, because these strains transmit the selectable phenotype with less than 100.% fidelity, it is impossible to estimate exactly the percentage of cells within the animal whichpossess the transgene. Basically in this analysis, the Iikelihaod of losing the array wasconsidered to be the same in all somatic lineages as in the germ line. This assumption may notbe true and certain cell lineages may preferentially retain the transgene while others may not.The number of copies did not seem to depend on the construct injected or even on thequality of each individual DNA preparation since all lines for one construct were produced frominjections with the same DNA preparation (Table 3).To determine if copy number correlated with the intensity of expression betweenstrains carrying the same construct, the induction of the transgenes over time was comparedwith the copy number (Table 4). Generally, copy number did not consistently correlate withthe first appearance or extent of IacZ expression. Even after relatively short heat shocks(30’), PC31, with only 14 copies per genome equivalent, stained as frequently in intestine andbody muscle as PC16 which contained 50 times the number of copies per genome equivalent. Inaddition this study indicated that the selection employed had no significant effect on the patternof expression. Both the twitcher and the rolling strains stained equally well in body muscle.69Chapter 1: ResultsPC4OPC38pc63 6 12 25genome pPD 16.43Figure 11. Dot Blot analysis of transgene copy number. Serial dilutions of genomic DNAs fromtransgenic strains were probed with radiolabelled primer extended IacZ. The copy number wasinferred from the linear region of a standard curve obtained using plasmid pPD1 6.43 whichcontains one complete IacZ gene per genome. Using the estimated value of 100 Mb for the size ofthe haploid genome of C. elegans, I estimated that one copy of the IacZ gene was equivalent toapproximately 1/33,000 of the haploid genome. Thus on the basis of loading 3 jig of genomicDNA the relative amount of pPD16.43 DNA required to represent a single copy gene wasestimated to be 90 pg. For the DNA dot blots, 3 jig of genomic DNA could not be effectivelyanalyzed because of the very high copy number in some strains. Samples were dilutedaccordingly and the results then extrapolated to represent 3 jig.ng DNA75II0.3 0.75 1.5copy no. per50 10070a>-.a>flCDkwrnmoo—C..,—lo•00000000000000C)000>[.c)..a>0>C.)r-Cncnc.c.-a>a>cu0)— Z- -C.)oN Ci)_8-000).A-4a>C.)I)a>0I)cu00rI’)-uDCI)3—1000CJ0(u000OQi0o(yiooo— (0•0 oCA>--.-<a>00Ci)0£30)0CflG)——m ZC.)oQ.mChapter 1: ResultsTable 4. Comparison of expression patterns between 1.48E1 strains.TRT. STRAIN COPY# SELEC11ON IN BM HYP PHA120’HS PC31 14 pRF4 8 10 5 230’REC PC33 33 pRF4 9 10 8 6120HS PC16 750 pPD1O.41 7 9 10 6120’HS PC31 10 10 6 10PC33 9 9 8 5PC16 7 9 9 990’HS PC31 9 8 5 7PC33 9 10 10 10PC16 6 9 7 .860’HS PC31 7 6 3 1PC33 8 10 4 5PC16 7 9 8 830HS PC31 9 9 4 4PC33 4 10 5 1PC16 7 8 8 515’HS PC31 10 4 0 0PC33 9 6 5 3PC16 1 10 8 4For each datum point ten staining animals were scored for expression in the specified tissue.HS, Heat shock; REC, Recovery; IN, Intestine; BM, Body muscle; HYP, Hypodermis; PHA,Ph ary n x.72Chapter 1: ResultsHypodermal expression was somewhat reduced in one of the rolling strains, PC31, but not in theother, PC33.73Chapter 1: Discussion0: DISCUSSION1. The lacZ transgenes are correctly expressed in response to heat shock in vivoThe j. localization studies described here, as well as previous studies of hspl 6mRNA levels (Russnak et aL,1985; Jones et al., 1986) demonstrate that these genes arestrictly heat inducible and are not expressed constitutively at any time in development. All ofthe hspl6-lacZ transgenes in this study were expressed only in response to a heat shock.Neither the copy number (as high as 750 copies per cell in some strains) nor the complexnature of the extrachromosomal arrays seemed to interfere with the tight heat inducibility ofthese promoters.2. Transgene copy number versus expression levelsExamination of multiple strains carrying each construct showed no consistentcorrelation between gene copy number and levels of expression. The primer extended lacZproducts used to probe DNA should hybridize to any lacZ sequence of reasonable homology andlength; whether it be a complete contiguous coding region or not, the probe cannot distinguishbetween these events. In light of the fact that different sized fragments of the selection and testplasmid sequences are presumably mixed in head to head and head to tail conformations, theactual number of complete transcribable copies of the lacZ transgene is probably much lowerthan dot blot analysis suggests. On the other hand strains which contain high copy numbers•(e.g. PCi 6 with 750 copies) substantially increase the number of HSEs per genome. Thus, evenif all copies of the transgene are not complete, a good deal of HSF could be sequestered at theseHSEs during stress. It is conceivable that levels of HSF could be limiting and that strains with750 transgene copies (or 1500 additional HSEs) may have surpassed that limit such that manyHSEs remain unbound by HSF during stress.In addition the presence of mixed arrays heightens the possibility of positional effectsdue to sequences derived from the selection plasmid or vector sequences. There is evidencewith other genes which have been co-injected with the unc-22 plasmid pPD1O.41, that74Chapter I: Discussionenhancer sequences can promote spurious expression in muscle cells (Andrew Fire, personalcommunication). Since muscle expression was strong in strains transformed with the rollerconstruct, pRF4, and since variability in muscle expression was between strains transformedwith different fusions rather than between strains transformed with identical fusions, I believethat any such effects are minimal or non-existent in this analysis. Other studies havesuggested that the roller plasmid can have negative effects on expression. This effect isparticularly enhanced in late larval stages and in gene fusions which contain short promotersequences (Andrew Fire, personal communication). Thus there is a possibility that the limitedexpression observed in the transcriptional fusions may in part reflect negative interactions.3. Expression of the hspl6 gene pairs is tissue generalThe staining patterns observed for the whole locus lacZ fusions (48.1C, 41.2C) verylikely reflect the in viva situation since all hspl6 coding and flanking sequences are includedin these constructs. In spite of the quantitative differences in expression between loci, asillustrated by the exon 1 fusions, most cells transcribe mRNAs from all four hspl6 genes.Moreover, all four genes first become heat inducible in gastrulating embryos, and are inducibleat all subsequent stages of C. elegans development. Only the germ line and the early embryo failto express the hspl6 genes following a heat shock. Thus the HSP16 proteins likely provide afunction which is required by all cells subjected to heat shock. Like their Drosophilacounterparts and the a-crystallins, the HSP1 6s of C. elegans form large aggregates in viva(Hockertz et aL,1991). Speculation concerning the functions of these proteins has includedroles in protecting mRNA (Nover et aL, 1989), the cytoskeleton (Leicht et al., 1986; Vierlinget al., 1988) or organelles (Cooper and Ho 1987) from the effects of heat shock; however, nodefinitive role for the small HSPs has yet been established.Curiously, the nuclear localization signal of 41 .2C failed to direct the fusion protein tothe nucleus, although this occurred efficiently in 48.1C transformants. This implies that thetwo fusion proteins fold differently in viva such that the NLS was exposed in the 48.1 C gene75Chapter I: Discussionproduct but not in the 41 .2C gene product. This was surprising since the restriction site usedto insert IacZ into hspl6-1 and hspl6-2 was a Hpa I site present in the identical position inboth genes. In addition the integrity of the NLS in 41 .2C was verified by restriction digestionand sequence analysis. The HSP16-1 and HSP16-2 proteins are 90% identical throughouttheir sequences (Jones et aL, 1986) and the nearest amino changes relative to the lacZ proteininsertion are 9 amino acid residues away on the N-terminal side, and 26 on the C-terminalside. It is possible that these few amino acid differences can significantly alter the local foldingand consequently perhaps the biological properties of these HSP16s. Alternatively, the N-terminal region of the 41.2C fusion may be unstable due to proteolysis, or some otherunidentified signal within the hspl6-2 coding region may interfere with the function of theNLS. Possibly relevant to this question is the fact that construction of the 48.1C fusion resultedin insertion of an extra 8 codons upstream relative to 41 .2C. This may have served as a spacerin 48.1C, allowing better function of the NLS.4. Tissue differences in the expression of hspl6-lacZ transgenesThe quantitative differences in expression observed between the hspl6-48/1 andhsp 16-4 1/2 gene pairs, as illustrated by the exonl -lacZ gene fusions, may also reflectdifferent cellular requirements if in fact different HSPs possess slightly different functions.The intergenic regions of these pairs are 86% identical. It is not yet clear what sequencesaccount for the differences in tissue expression. These differences could result from eithertranscriptional or post-transcriptional effects; the latter might be mediated, for instance,through differences in the mRNA leaders. For example, sequences in the leaders of theDrosophila hsp7O and hsp22 mRNAs have been shown to be required for selective translation ofthese genes under heat shock conditions (McGarry and Lindquist, 1985; Hultmark et aL, 1986;). Previous evidence suggests that the hspl6 pairs are differentially regulated (Jones et al.,1989; Dixon eta!., 1990). Up to 14-fold more RNA transcripts are present per gene from thehspl6-41/2 pair relative to the hspl6-48/1 pair after a two hour heat shock. Moreover, it76Chapter I: Discussionwas found that while hspl6-2 mRNA was more abundant than that of hspl6-1 at all points inthe life cycle, the difference was most pronounced for embryo and Li stages (Jones et aL,1989). No differences in the magnitude of expression between the 48.1C and the 41.2Ctransgenes were observed in this study. However, since staining intensity was assessed visuallyand since the SV4O NLS was non-functional in the 41 .2C transgenic animals, differences couldeasily have gone undetected.The finding of intra-pair differences in expression of the hspl6-48/1 transgenesimplies the existence of a directionality to the intergenic region. It remains undeterminedwhich sequences are involved in this phenomenon. Transgenic strains carrying 48.1XBE1, aconstruct which inverts the rgion of the 48.1E1 transgene between the hspl6-1 and hspl6-48 HSEs failed to significantly alter the pattern of expression toward that of 1 .48 El transgenicanimals. The leader sequences, however, remained unchanged in this experiment and couldaccount for this result. Interestingly, neither 48.1E1 or l.48E1 transgenic worms stained aswell in neural tissue as the whole locus fusion, 48.1C transformants. This may be the result ofdifferences in the stability of the RNA produced from these constructs. In particular, the wholelocus fusions have retained the endogenous hsp-16 3’ sequences whereas the exon 1 fusionsutilized a 3’ end from a body wall myosin gene, unc-54 ( Epstein et aI.,1974; Fire et a!.,1990). It is possible that the 3’ ends of the hspl6 genes may be important for mRNA stabilityduring heat shock.5. Cooperative interaction of HSEs may enhance expressionElimination of the HSEs and TATA box for one gene of a pair in the 48P and 41 Pconstructs significantly reduced the production of -galactosidase in somatic tissues, such thatdouble heat shocks were required to assess somatic expression, especially in 48P transgenicanimals. The 48P promoter was approximately 70 bp shorter than the 4lP promoter andlacked the transcriptional start site of the endogenous gene, which may account for the weakerand more restricted expression observed in strains carrying the former construct.77Chapter I: DiscussionNonetheless, production of f-galactosidase in somatic tissues in the fusion 41P wassignificantly reduced in comparison to the fusion 2.41 El. Essentially two differences exist inthe construction of these fusions. The hspl6-41 breakpoint proximal to IacZ in the 41P fusionis two bp upstream of the hspl6-41 initial methionine; thus it is a transcriptional fusion. Thedistal breakpoint is located within the hsplô-2 HSE such that the element is destroyed in the41 P construction. Thus while 41 P contains all of the sequences between the HSEs as well as thetranscriptional start site, HSE and TATA box of the hspl6-41 promoter, the HSE and TATA boxof the hspl6-2 promoter have been eliminated. The 2.41E1 fusion on the other hand was atranslational fusion of the first 15 amino acids of hspl6-41 to IacZ and possessed the HSEs andTATA boxes of both genes. The reduction in expression of the 41 P fusion therefore may be theresult of losing the translational start site of the hspl6-41 gene; a less likely possibility isthat the closer proximity of vector sequences in 41 P may inhibit expression. I believe that themost likely explanation for the observed difference in expression is the existence ofcooperativity between the two hspl6 promoters in the wild type gene pairs. This is inagreement with the results obtained when the hspl6 genes were introduced into mouse cells(Kay et a!., 1986). Recent evidence suggests that in vivo , HSTF functions as a trimer orperhaps even a hexamer (Sorger and Nelson 1989; Cbs et at, 1990) and binds to HSEs atrepeating nGAAn motifs, which may be in either orientation (Amin et a!., 1988; Xiao and Lis1988; Perisic et at, 1989). The HSEs of each hspl6 gene at each locus possess three suchsequences. Thus each gene promoter is capable of binding an HSTF trimer.It has been suggested that HSTF trimers binding at separate HSEs can interact with eachother cooperatively to enhance transcription (Shuey and Parker 1986; Xiao et at, 1991). TheHSEs at each hspl6 locus are only 130 bp apart, suggesting that such a cooperative interactionmay be possible. It has been shown that HSEs positioned as much as 2 kb apart can activatetranscription cooperatively (Riddihough and Pelham 1986), and Thomas and Elgin (1988)have shown that HSEs separated by more than 200 base pairs in the Drosophila hsp26 gene maybe brought together via folding around a nucleosome core positioned over the intervening region.78Chapter 1: DiscussionThus a divergent arrangement of closely spaced hspl6 genes may lead to more efficient heatinducible expression for a given number of heat shock elements.While temperatures of 32° to 33°C are required for consistent somatic tissue expression,29°C is sufficient over the same time period for embryos. Furthermore, transcriptionalfusions were expressed extremely well in embryos while expression in the somatic tissues oflater stages was minimal. These results suggest that the minimal requirements for high levelexpression in embryogenesis are maintained in these fusions while some element(s) necessaryfor somatic expression have been lost. Alternatively, it is possible that embryos have largerlevels of available heat shock transcription factor (HSTF) relative to the somatic tissues oflater stages. If so, then somatic gene expression may rely more heavily on the cooperativebinding of HSTF to the HSEs in the intergenic region to achieve a given level of expression.6. Developmental Regulation of the hspl6s in C elegansAlthough the hspl6 genes are apparently never expressed constitutively, several lines ofevidence indicate that their inducibility is regulated at certain stages in development. Firstly,no expression has been observed in the germ line of any of the transgenic strains studied todate. Secondly, embryonic expression was not consistently observed until gastrulation. It hasbeen suggested that in C. elegans, zygotic genes are first transcribed at gastrulation (Hecht eta!., 1981) and that this stage loosely corresponds to the mid-blastula transition in amphibians(Newport and Kirschner 1982). The results presented in this study suggest that adevelopmentally regulated program overides the heat inducibility of the hspl6 genes: thematernal genes are not inducible during gametogenesis or early embryogenesis, while thezygotic genes become responsive to heat shock at approximately the time of onset of generalzygotic transcription. With respect to gonadal expression however, it is notable that othertransgenes (such as msp-1 which encodes a sperm specific product), fail to be expressed in thenematode gonad. This suggests that there may be a general barrier to transgene expression inthis tissue (D. Dixon, personal communication).79Chapter 1: DiscussionIt is interesting to compare the regulation of the hspl6 genes to that of the related genefamily in Drosophila . The small hsps of Drosophila, hsp22, 23, 26 and hsp27 are alsodevelopmentally regulated but expression is induced by the hormone ecdysterone duringpupation and oogenesis in the absence of heat shock (Ireland and Berger 1982). Ecdysteroneinduction is mediated through an element which is separate and distinct from the HSEs requiredfor heat induction (Cohen and Meselson 1985; Hoffman and Corces 1986). To date four hspl6genes from C. elegans have been cloned and seem to be purely heat inducible, It is possible,however, that there may be related members of this family which are expressed temporally inthe absence of heat shock. Recently, a gene encoding an 18 kD protein which is cadmiuminducible has been cloned and found to share sequence similarity to the HSP16s (C. Rubin,personal communication). Exposures to cadmium which are sufficient to activate this gene donot consistently induce the hspl6 transgenes used in this work, but in one experiment intensestaining in the pharynx, intesine and vulva of some animals (approx. 1 %) was observed.In Xenopus a temporal pattern of expression is superimposed upon the heat inducibilityof some of the HSPs. For instance, hsp3o first becomes heat inducible at the tadpole stage, andthe levels of mRNAS of this gene vary considerably from tissue to tissue in adult frogs, beinggreatest in the kidney and gut (Bienz 1984). Similar phenomena were observed in this study,but the quantitative differences in spatial expression were seen among members of the samegene family.7. Conclusions1) The hspl6-lacZ transgenes were correctly regulated with respect to heat shock intransgenic nematodes. Expression was entirely heat dependent in aggreement with previousresults obtained by Russnak et al. (1985) and Jones et al. (1986). Heat shock resulted inrapid and intense induction of f-galactosidase activity whereas no staining was observed inanimals maintained at normal growth temperatures. These results demonstrate the validity ofemploying extrachromosomally inherited transgenes to study gene expression in C. elegans.80Chapter 1: Discussion2) Inducibility of hsplB genes appears to be under developmental regulation since expressionwas not inducible in the germ line or early embryogenesis for any of the transgenes studied.3) Expression of the hspl6 gene pairs is temporally and spatially non-specific fromgastrulation onwards.4) Sequences conferring tissue specificity are contained within the intergenic region of thehsplG gene pairs since orientation of this region affects the priority of tissue expression.5) The experiments described in this study represent the first fully inducible system to becharacterized molecularly in C. elegans and offers unique opportunities to the C. elegansinvestigator who wishes to employ a reverse genetic approach.8. Future ProspectsThis study pursued a reverse genetic approach to tackle analysis of the hspl6s as theyare expressed in C. elegans. In light of the results obtained, a number of interesting avenuescould be followed to elucidate further the role of the small HSPs in the stress response. Anantisense strategy could be employed to approximate loss of function alleles in order todetermine if the HSP1 6s are essential during stress or for specific components of the stressreponse. Such experiments may provide extra information regarding the identity of theHSP1 8s.In addition, the differential expression revealed by the exon 1 fusions suggests thatmultiple enhancing elements may be present within the intergenic region. Thus, furtherdissection of this region would be desirable to identify these sequences. Ultimately, tissuespecific enhancers may be identified which could be employed to direct expression of any gene ofinterest to specific tissues.The strict heat inducibility of the hspl6 genes constitutes a potentially powerful toolwhich could be used to address a variety of interesting biological questions in C. elegans. Usingthese promoters, it should be possible to achieve the heat inducible expression of almost anydesired coding region. Since the hspl6 promoters can be activated to produce a high level of81Chapter 1: Discussionexpression within minutes, it is feasible to examine the effects of a particular gene productwhen produced at short, specific times in development. For example, an hspl6-mab5 fusionhas been used to achieve tightly controlled, heat-dependent movement of specific migratoryneuroblasts at the Li larval stage in a mab-5 null background (S. Salser and C. Kenyon,personal communication).By combining tissue-specific enhancers with heat shock promoters, it should bepossible to achieve heat inducible expression in specific target tissues, as suggested by Peiham(1987). Such a hybrid promoter, fused to a gene which produces a toxic product such asdiphtheria toxin, might allow specific killing of selected cells under controlled conditions. Inanother approach, it may be possible to use a laser (White and Horvitz, 1979) to heat shockindividual cells within an animal containing an hsplô gene fusion of interest. The transparencyof C. elegans and its defined cell lineage (Sulston et aL, 1983) make it particularly well suitedto such an application. Such a technique, if feasible, would provide unique opportunities toassess the functions of particular gene products in chosen cells at a given point in time, andthereby to study the roles of such products in cellular interactions during determination anddevelopment.The induction of HSPS by a variety of biologically harmful agents including heat shockhas prompted speculation that these proteins could be used to monitor environmental stress.One of the problems with this approach is that many HSPs are produced at low levelsconstitutively or at specific times in development in the absence of stress, thus assays wouldhave to detect increases above background levels. The hspl6 genes described in this study areonly expressed in response to heat shock but future experiments will include determining ifthese genes are induced in response to chronic stress by a variety of agents. If these results areaffirmative then the hspl6-IacZ transgenes described could be powerful tools in theassessment of environmental stress. Firstly, C. elegans is a natural inhabitant of soil;secondly, the histochemical staining procedure is a simple assay for expression which requiresminimal technical dexterity and has a clear positive or negative result. Moreover, the assay is82Chapter 1: Discussionsensitive, and HSP induction does not result in lethality. Finally, in conjunction withdevelopmental time curves, these animals could provide invaluable information regarding theeffects of chronic non-lethal stress on development and ageing. On the practical side, however,these experiments will necessitate the construction of homozygous integrated lines which can beeasily maintained in the absence of selection.Recently, using an X-ray mutagenesis procedure suggested by Jeff Way (personalcommunication), I generated stably integrated homozygous rolling lines from a 48.1Cextrachromosomally transmitted strain, PC6. Initial analysis of one of these strains, PC71,substantiates the findings presented in this thesis. The IacZ transgene is expressed only inresponse to heat shock and in the tissue general manner characteristic of its ancestor, PC6(48.1C/pRF4). Rarely, a few (< 5 %) comma to pretzel staged embryos also stain, perhapsdue to positional effects. Future endeavours will include mapping experiments to determine thesite of array integration, the copy number and integrity of the transgene. Since the X-raymutagenesis procedure was. technically simple and effective it should be relatively trivial togenerate homozygous lines for all constructs of interest, bypassing the tediousness of injections.This would make biochemical experiments a more accessible prospect, as well as enhance thepotential for using these animals as probes of environmental stress.83Chapter II: IntroductionII:ANALYSIS OF POLVUBIQUITIN GENE EXPRESSIONA: INTRODUCTION1. General IntroductionUbiquitin is a highly conserved 76 residue polypeptide which is present in alleukaryotic cells (Goldstein et al., 1975). Determination of the structure of ubiquitin at 0.28nm and later at 0.18 nm resolution revealed that it is a highly compact globular protein withan amino terminus buried within a hydrophobic core and a protruding carboxy terminal tail(Vijay-Kumar et at., 1985; Vijay-Kumar et al., l987a; Vijay-Kumar et al., 1987b).Ubiquiti.n is highly resistant to thermal and chemical denaturation (Ciechanover et at., 1978;Lenkinsky et at., 1977).Numerous cellular processes are mediated by the post translational conjugation ofubiquitin to proteins via an isopeptide bond between the carboxyl terminal glycine of ubiquitinand the epsilon amino group of an internal lysine residue in the target protein (see Hershko andCiechanover, 1986; and Hershko, 1988 for reviews). In this fashion, it is believed thatubiquitin “marks” a protein for a particular fate: whether it be for degradation by the nonlysosomal ATP dependent pathway, or for some other protein interaction.2. Ubiquitin mediated proteolysisIn the cytoplasm, ubiquitin is involved in the ATP dependent proteolysis of damaged ordefective proteins (reviewed in: Hershko and Ciechanover, 1986; Hershko, 1988). The eventsinvolved in this procedure are summarized briefly in Figure 12 and in the text below. First,ubiquitin is activated in a two-step reaction by a specific enzyme, El (Step 1 of Figure 12).Ubiquitin adenylate is formed with the displacement of pyrophosphate from ATP. Subsequently,activated ubiquitin is transferred to a thiol site on the El enzyme and AMP is released.In step 2, ubiquitin is transacylated from El to a specific cysteine residue on a carrierenzyme termed E2. Many different E2 enzymes have been identified, some of which can catalyze84Chapter II: IntroductionFig. 12. The non-lysosomal ATP dependent proteolytic pathway. Adapted from Hershko(1988). See text for details. Ub, ubiquitin; El, ubiquitin activating enzyme; E2, ubiquitinconjugating enzyme; E3, ubiquitin ligase; PCFs, protease complex factors.PeptidesAMP+PPIATPUb-cOH8ProteasetATPComplex \.4\•7ATPI 6PCFS0El -S-C-Ub2 E2-SHsHistones“3E2-S-C-Ub -b- Ub-Hstone5 ProteinE3-Protein E3(U b) n - P rote inConjugates85Chapter II: introductionthe assembly of stable ubiquitin-protein conjugates which are not destined for degradation(Step 3 and Section 5).Proteins with structures suitable for degradation are selected by and complexed toubiquitin ligase or E3 enzymes (Step 4). One of these E3 enzymes, that encoded by the UBR1gene of yeast, has been shown to specifically recognize and bind only to proteins which have destabilizing amino terminal residues (Bartel et al., 1990) as defined by the N-end rule(Bachmair et al., 1986).In step 5, ubiquitin is transferred from the E2 complex and linked by its COOH glycine 76residue to E-NH2 groups of internal lysine residues in the E3 bound target protein. In this way,numerous monomeric ubiquitin molecules can be bound to several lysine residues in a protein,and also form branched polyubiquitin chains stemming from a single lysine. The latterformation has been implicated as a pre-requisite for degradation (Chau et al., 1989).Protein-ubiquitin conjugates are degraded to small peptides by a large (> 1 OOkDa)protease complex (Steps 6 and 7) the formation and action of which is dependent on ATP.Finally, in the last step of the model, free ubiquitin is recycled for use in subsequentdegradations (Hershko, 1988).3. Ubiquitin as a regulatory proteinRecently it has been demonstrated that a number of short-lived proteins are alsodegraded by the ubiquitin-dependent system. This includes the nuclear oncoproteins encoded byN-myc, c-myc, c-los, and EM (Ciechanover et al., 1991), and the yeast MA Ta2 repressor(Hochstrasser et al., 1991). Degradation of cyclin by the ubiquitin pathway is the key eventsignalling exit from mitosis (Glotzer et al., 1991). In plants, degradation of phytochrome isaccompanied by the formation of phytochrome-ubiquitin conjugates and is activated by light(Shanklin et aI., 1987). Thus by determining the half-lives of regulatory proteins, theubiquitin system may be an important regulator of cell cycle events.86Chapter II: Introduction4. Ubiquitin and heat shockA common result of physiological stress such as heat shock is the accumulation ofabnormal or damaged proteins. This presumably would place pressure on the ability of systemssuch as the ATP dependent proteolytic pathway to degrade and process the overload.Increased ubiquitinylation of proteins has been shown to accompany heat shock (Parag etal., 1987). A temperature sensitive mouse cell line, ts85, carries a mutation in an Elactivating enzyme (Finley et al., 1984) which severely impedes the formation of ubiquitinprotein conjugates under non-permissive conditions. As a result, these cells are unable todegrade efficiently short lived or abnormal proteins at restrictive temperatures (Ciechanoveret al., 1984). In addition, ts85 cells produce massive quantities of heat shock proteins atrestrictive temperatures compared to wild type cells (Finley et al., 1984; Ciechanover et al.,1984). The yeast genes UBC4 and UBC5 encode ubiquitin conjugating enzymes (E2s)which specifically conjugate ubiquitin to short lived or abnormal proteins destined forelimination (Seufert and Jentsch, 1990). These genes are also transcriptionally heatinducible. ubc4ubc5 double mutants accumulate free monomeric ubiquitin, fail to turn overproteins efficiently, constitulively express heat shock proteins at normal growth temperaturesand are inviable at elevated temperatures or in the presence of the amino acid analoguecanavanine (Seufert and Jentsch, 1990). This provides further evidence that the heat shockresponse is induced by the presence of excessive quantities of abnormal proteins. In retrospectthen, it is not surprising that ubiquitin itself is a heat shock protein in yeast, Drosophila, andchicken (Bond and Schlesinger, 1985; Finley et al., 1987; Lee et al., 1988). (See section 8.for more detail.)5. Ubiquitin and chromatinWithin the nucleus, histones are modified by ubiquitin conjugation in a reversibleprocess which does not trigger proteolytic degradation (Busch , 1984). It has been estimatedthat approximately 10% of histone H2A (West and Bonner, 1980a) and 1.5% of H2B is87Chapter II: Introductionubiquitinylated (West and Bonner, 1980b). Conjugation to histones H2A and H2B seems to be acharacterisitic of active chromatin (Levinger and Varshavsky, 1982; Nickel et al., 1989;Davie and Murphy, 1990).The RAD-6 gene (UBC2) of yeast encodes a ubiquitin conjugating enzyme (E2) whichtransfers ubiquitin to histones in an E3 independent reaction in vitro (Jentsch et al., 1987)and this may be related to its central role in DNA repair in vivo (Sung et al., 1988; Sung et al.,1990). Mutants of UBC2 are defective in DNA repair, mutagenesis and sporulation. Expressionof UBC2 is induced by ultraviolet light, and is required throughout the cell cycle but peaks inmeiosis during the period of maximal genetic recombination. Unlike UBC4 and UBC5,expression of UBC2 is not induced by heat shock or starvation suggesting that this E2 is notsubstantially involved in the conjugation of ubiquitin to damaged proteins during stress(Madura et al., 1990). However, UBC2 has been shown to be essential for multiubiquitinylation and degradation of N-end rule substrates which have been recognized by the E3ligase, UBR1 (Dohmen et al. 1991; Sung et al. 1991). Thus, this E2 catalyzes theformation of both stable ubiquitin-histone conjugates and unstable ubiquitin-protein conjugatesdestined for degradation.Conjugation of E2 enzymes to ubiquitin requires formation of a thioester adduct. Sitedirected mutagenesis of the sole cysteine residue in UBC2 totally abolishes ubiquitinconjugating activity and the biological function of UBC2 (Sung et al., 1990). This providesdirect evidence that ubiquitinylation of histones is required constitutively for DNA repair aswell as transcription.Another ubiquitin conjugating enzyme, UBC3 of yeast was originally isolated as a cellcycle mutation, CDC34. The protein product of UBC3 is required for the transition from the Gito the S phase in the cell cycle (Pringle and Hartwell, 1981) and catalyzes the conjugation ofobiquitin to H2A and H2B in vitro (Goebl et al. 1988). Thus ubiquitin-conjugation to histonesin chromatin may be an important regulator in a large number of nuclear events.88Chapter II: Introduction6. Ubiquitin at the cell surfaceVarious membrane proteins such as the lymphocyte homing receptor (Siegelman et al.,1986), the growth hormone receptor (Leung et al., 1987), and platelet-derived growth factorreceptor (Yarden et al., 1986) are conjugated to ubiquitin in vivo. Characterization of each ofthese receptors revealed two amino termini, one corresponding to the receptor polypeptide andthe other to ubiquitin. While the function of this arrangement remains unresolved, it has beensuggested that ubiquitin may act as a “tag” in cell guidance interactions during receptormaturation (Siegelman et al., 1986). Alternatively, ubiquitin may act in its conventional roleby mediating receptor turnover since the growth hormone receptor has been shown to be shortlived in vivo (Baxter et al., 1985; Gorin et al., 1985).7. Ubiquitin and myofibril assemblyIn Drosophila, the indirect flight muscles contain a myofibrillar protein called arthrin,which is a stable actin-ubiquitin conjugate (Ball et al., 1987). Ball et al. followed theincorporation of labelled [35S] methionine into actin and arthrin during muscle formation andshowed that arthrin synthesis lagged behind that of actin and that one arthrin molecule wasproduced per thin filament of muscle (1987). The authors suggested that arthrin may targetparticular actin monomers within filaments as appropriate or inappropriate binding sites formyosin heads during the cross-bridge cycle of myofibril assembly (1987).8. UBIQUITIN GENE STRUCTURE8.1 Polyubiquitin gene structureThe polyubiquitin gene consists of tandem repeats of a ubiquitin coding region which istranscribed as a polycistronic mRNA, translated into a polyprotein and cleaved posttranslatiorially to release ubiquitin monomers (Lund et al., 1985; Ozkaynak et al., 1984). Nospacers or introns separate the ubiquitin repeats; the carboxyl-terminus of one ubiquitin89Chapter II: Introductionmoiety is linked to the amino-terminus of the next in the polyprotein while the last ubiquitinrepeat carries a one to three amino acid extension (Table 5).The number of polyubiquitin genes per genome and the number of ubiquitin codingrepeats per polyubiquitin gene varies considerably between organisms (Table 5). The yeastpolyubiquitiri gene, UBI4, contains five ubiquitin coding regions in a head to tail arrangement(Ozkaynak et al., 1984); Drosophila polyubiquitin contains 18 repeats (Lee et al., 1988), andDictyostelium contains two genes of three and five repeats respectively (Giorda and Ennis,1987). In spite of this variation in ubiquitin repeat number, the amino acid sequence ofubiquitin monomer is largely invariant from species to species: for example, yeast and humanubiquitin differ by only three residues (Table 5 ). Thus, ubiquitin structure and function arelargely intolerant of sequence change.8.2 Polyubiquitin gene expressionIn yeast the single polyubiquitin gene, UBI4, is essential for sporulation, resistance tostarvation and heat shock (Finley et al., 1987), and GO /Gi arrest (Tanaka et al., 1988).Deletion mutants of UBI4 are viable under normal conditions of exponential growth, andmaintain wild type levels of free ubiquitin which is presumably derived from the ubiquitinribosomal fusions (Finley et al., 1987; Ozkaynak et al., 1987; See section 8.3). However,under conditions of physiological stress, the fusion proteins apparently cannot maintainsufficient quantities of ubiquitin monomer in UBI4 mutants (Finley et al., 1987) and thesecells are hypersensitive to heat and are sporulation defective. These defects werecomplemented by the introduction of a UBI4 minigene which contained the UBI4 5’ upstreamsequences and one ubiquitin coding repeat. Thus, UBI4 gene function is to provide ubiquitinmonomers during stress as opposed to polyprotein (Finley et al., 1987). While UBI4represents the sole yeast polyubiquitin gene, two mRNA transcripts of 2.6 kb and 1.5 kb wereobserved on Northern blots (Finley et al., 1987) and the levels of the 1.5 kb transcript wereelevated upon heat shock. Thus, the UBI4 gene is induced by heat shock at the transcriptional90Chapter II: IntroductionTable 5. Comparison of polyubiquitin genes among several species. The conservation ofamino acid sequence amongst species is shown as well as the variation in gene number, repeatnumber and carboxyl terminus extension.Chicken 0yeast 32Drosophila 0C. elegans 1Tetrahymena 2Baker andBoard, 1987;Wiborg et al.,1 9853; 4 Tyr Bond andSchlesinger,1 985Asn Ozkaynak etal., 1984Leu; Asn Giorda andEnnis, 19871 8 lIe GIn Ala Lee et al.,1 988Asp lieal., 19891?, 3?,... Ser; GIn Neves et al.,1 988Organism # residues # # ubiquitin carboxy Referencedivergent polyubiquitin repeats! gene terminalfrom human genes extensionsequenceHuman- 2 3; 9 Cys; ValDictyostelium521>=611>=43, 5, ?...11 Graham et91Chapter II: Introductionlevel, and sequences matching the heat shock element consensus have been identified upstream ofthe UB!4 coding region (Ozkaynak et at., 1987). Multiple RNA species transcribed from asingle polyubiquitin gene have also been described in Xenopus (Dworkin-Rastl et al., 1984).Neves et al. (1988) showed that Tetrahymena possess at least four polyubiquitin geneswhich are differentially expressed during heat shock. Three different RNA species (5.6 kb, 1.8kb, and 0.75 kb) were observed on Northern blots from unstressed cells whereas heat shockincreased the levels of the 5.6 and 1.8 kb transcripts and induced the appearance of a 1.6 kbtranscript (Neves et at., 1988). Expression of the two polyubiquitin genes of chicken, Ubi andUbil, is also differentially regulated. Levels of the two polyubiquitin RNAs were determined inthree tissues- mature testis, immature testis, and liver reticulocytes. While the Ubitranscript was the most abundant in all tissues, the proportion of transcripts derived from theUbli gene increased from 8% in immature testis to 22% in mature testis. In addition, anoverall increase in ubiquitin levels was observed during spermatogenesis suggesting that theincreased demand for ubiquitin during spermatogenesis is met by increased transcription of theUbil gene (Rocamora and Agell, 1990). While the Ub!! gene seems to be important forspermatogenesis, transcription of Ubl was induced by heat shock and chemical stress (Rocamoraand Agell, 1990), and in fact Ubi was initially identified in a screen for heat shock genes(Bond and Schlesinger, 1985).Drosophila contains one polyubiquitin gene Of 18 tandem ubiquitin repeats (Lee et at.,1988). Expression of this gene is constitutive but appears to be moderately heat inducible. Leeet al. (1988) reported a three-fold increase in polyubiquitin expression during heat shock.Using P element mediated transformation of a polyubiquitin-lacZ fusion, they determined thatthe polyubiquitin gene was expressed constitutively in a developmentally non-specific andtissue general manner (Lee et at. 1988).Based on the above data, a few generalizations with regard to polyubiquitin geneexpression can be made. In organisms such as Drosophila, where there is only onepolyubiquitin gene, expression of this gene seems to be temporally and spatially non-specific92Chapter II: Introductionbut moderately inducible by physiological stress. Heat induction can, as in yeast, result inmultiple RNA species being transcribed from the same gene. Organisms such as chicken, whichcontain more than one polyubiquitin gene, show differential expression amongst these geneswith respect to development and physiological stress.8.3 Ubiquitin Fusion genesUbiquitin fusion genes consist of a ubiquitin repeat fused to one of two basic “tail”sequences which encode small ribosomal proteins (Finley et al. 1989; Ozkaynak et al. 1987;Lee et at., 1988; Redman and Rechsteiner, 1989). Thus, unlike the ubiquitin-proteinconjugates which are formed post translationally by the action of ubiquitin conjugatingenzymes, these gene fusions are translated into a linear bi-protein molecule which is cleavedpost-translationally to release monomeric ubiquitin and the ribosomal protein in a manneranalogous to the processing of polyubiquitin. It has been proposed that fusion with ubiquitin maystabilise these ribosomal proteins and/or aid in ribosome assembly (Finley et at. 1989)..Such fusion genes have been found in a wide variety of eukaryotes including yeast,Drosophila and mammals (Finley et al., 1989; Redman and Rechsteiner, 1989;Lee et al.,1988), and exhibit a high degree of similarity in the tail sequences. For example, theriboscmal protein tail of the Drosophila UB3-D fusion is 65% identical to its yeast homologueand 82% identical to the human fusion. Moreover, Southern analysis in yeast indicates thatthese ribosomal proteins are encoded only by fusion genes (Otaka et al., 1984). Thus it islikely that these proteins provide a basic function required by all eukaryotes.Three of the four ubiquitin genes of yeast, UBI1, UBI2, and UBI3 encode fusion proteins(Finley et al., 1989). Deletion of any of these genes results in a slow growth phenotypewhereas deletion of UBI4 has no effect (Finley et at., 1989). Stow growth is a consequence ofloss of the fusion tail proteins since levels of free ubiquitin are almost normal in these cellsand since transformation of a plasmid carrying the tail sequence atone can complement thedefects (Finley et al., 1989). The tail sequences of the UBI1 and UBI2 genes encode the identical93Chapter II: Introduction52 amino acid protein which has been identified as a constituent of the large (60S) ribosomalsubunit. UBI1/UBI2 double mutants are inviable: thus these genes are essential in at least asingle copy (Finley et al., 1989). The UBI3 tail protein consists of 76 residues and is acomponent of the small (40S) ribosomal subunit (Finley et aL, 1989).While ubiquitin fusion genes are apparently necessary for wild type growth, they arenot induced by physiological stress as is often the case with their polyubiquitin counterpart(Ozkaynak et al., 1987; Finley et al., 1987; Finley et al., 1989).8.4 Ubiquitin like genesRecently, several genes have been identified which show sequence similarity to ubiquitinbut which do not encode products which function as conventional ubiquitin does (i.e. by ligationto other proteins). An interferon-inducible 15 kDa protein discovered in mammalian cellscontains two related domains, each with approximately 30% amino acid identity to humanubiquitin (Haas et al., 1987). The human GdX gene contains an amino terminus of 74 aminoacids bearing 43% identity to human ubiquitin (Toniolo et al., 1988).9.0 Trans-splicingIn trypanosomes, all pre-mRNAs receive a 39 nucleotide leader RNA molecule through atrans-splicing reaction (Laird, 1989). Trypanosome genes contain no introns and thisorganism does not possess Ui small nuclear RNA (5nRNA), the component of the spliceosomewhich recognizes the 5’ splice site in conventional cis-splicing. In trans-splicing the splicedleader (SL) RNA provides the 5’ splice site for the reaction and exists in viva as an RNPparticle (Van Doren et al., 1988; Bruzik et al., 1988; Thomas et al., 1988).For years trans-splicing was considered to be a phenomenon unique to trypanosomes,being an adaptation for efficient processing of the polycistronic messages common in theseprotozoans, and presumably incompatible with the cis-splicing of higher eukaryotes. Thediscovery by Krause and Hirsh (1987) that three of the tour actin genes of C. elegans94Chapter II: Introductiontranscribe pre-mRNAs which acquire a 22 nucleotide leader via trans-splicing shattered thisconception. It is now apparent that approximately 15 % of C. elegans transcripts are transspliced (Blumenthal and Thomas, 1988), including ubq-1 hnRNA (Graham et al. 1988).Moreover, while in trypanosomes, all RNA molecules acquire the identical 39 nucleotide splicedleader, in C. elegans two different spliced leader molecules (SL1 and SL2) have been identified(Huang and Hirsh, 1989) and each leader is spliced to specific nematode transcripts. Inaddition, in C. elegans trans-splicing and cis-splicing can occur not only within the samenucleus but within the same transcript (Krause and Hirsh, 1987; Graham et al., 1988). IndeedC. elegans possesses Ui snRNA while trypanosomes do not (Van Doren and Hirsh, 1988).Many similarities exist between cis and trans-splicing. Both cis- and trans-splicinginvolve the formation of branched intermediates: a lariat structure in cis-splicing (Grabowskiet al., 1984; Padgett et al., 1984; Ruskin et al., 1984), versus a Y-shaped forked molecule intrans-splicing (Sutton and Boothroyd, 1988). The 5’ splice site sequence of SL1 RNAresembles typical 5’ splice sites of C. elegans introns, while the 3’ acceptor site for SL1 or SL2RNA is virtually identical to a typical 3’ splice site in cis-spliced introns (Blumenthal andThomas, 1988). Recently, Conrad et al. (1991) demonstrated that inserting a normally cisspliced intron, devoid only of its 5’ splice site, upstream of the coding region of a vit-2/ vit-6fusion gene converted the non trans-spliced transgene into one which was efficiently transspliced to SL1. These results suggested that trans-splicing by SL1 is a default mechanismwhich occurs whenever a 3’ splice site is present unaccompanied by a 5’ splice site partner.What remains undetermined is what specifies a transcript for splicing by SL2.10. The present studyIn Caenorhabditis elegans, the polyubiquitin gene ubq-1 codes for eleven tandem repeats ofubiquitin (Graham et al. 1989; Figure 13). The upstream region of ubq-1 contains severalfeatures which might be involved in gene regulation, including a cytosinerich block, aninverted repeat possessing homology to the DNA binding site of the mammalian steroid hormone95Chapter II: IntroductionRNA STARTI >. ATS3ss--—- -II4II1 I II1-6uBQ-1I II 11 It 4 *t1.hnRNASLJ RNAmRNAFig. 13. Organization, transcription and processing of the polyubiquitin gene, ubq-1, ofCaenorhabditis elegans. Adapted from Graham et al. (1989). Ubq- 1 encodes 11 repeats ofubiquitin. The transcriptional start site is indicated by the bent arrow. 3’ss, acceptor site forsliced leader RNA; SL1 RNA, splice leader 1 RNA.96Chapter Ii: Introductionreceptor, and two sequences resembling heat shock elements. Ubq-1 is unique amongpolyubiquitin genes isolated to date in that its heteronuclear RNA acquires a 22 nucleotide leader(SL1) by a trans-splicing reaction (Graham et al., 1988). Also unique is the presence of cisspliced introns within the coding region (repeats 1, 4, 7, and 10) (Graham et al. 1989).Northern analysis shows that ubq-1 is expressed constitutively at equivalent levels throughoutdevelopment, and is not subject to regulation by heat shock or nutritional deprivation (Grahamet al. 1989).The purpose of this study was to examine the spatial expression patterns of ubq-1 intransgenic nematodes carrying ubq-1-IacZ fusions. In addition, when this study was firstconceived, no information regarding the specificity and mechanism of SL1 trans-splicing topre-mRNAs in the nematode was available. Thus it was hoped that transgenic studies of ubq-1might provide further elucidation of this phenomenon.97Chapter II: MethodsB: METHODS1. Maintenance of strainsNematode strains were maintained as described in Chapter I methods.2. Construction of ubq-1-IacZ fusionsUbq938-lacZ contains 1044 bp of ubq-1 sequences extending from a Sal I site 938 bpupstream of the initial methionine to an Eco RI site 103 bp downstream of the methioninecloned into the IacZ expression vector, pPD16.43 (Fire et al. 1990). Thus Ubq938-IacZ andall of the deletions are in-frame translational fusions containing the first 36 amino acids of thefirst exon of ubq-1. Deletions extending from the Sal I site of ubq938-IacZ, generated byexonuclease Ill digestion (Henikoff 1984) and analysed by sequencing ( Sanger et al. 1977),were the gift of Don Jones.A 1033bp Sal Ito Hind Ill fragment of ubq-1 3’ sequence was inserted between the Eag Iand Apa I sites of pPD1 6.43 (Don Jones, personal communication). The resulting constructcarried an additional 74 bp of polylinker derived from cloning vectors upstream of the start ofthe ubq-1 3’ non-coding sequence and an additional 774 bp of ubq-1 3’ sequence notpreviously published. The ubq- 1 polyadenylation signal is 110 bp downstream from the ubq1 stop codon and 154 bp downstream of the Sal I site.3. Establishment of transgenic strainsTransgenic strains were established in methods described in Chapter I. For transientassays the IacZ fusion alone was injected at a concentration of 200 ng/p.l.4. Identification of /3-galactosidase staining cellsAs described in Chapter I.98Chapter II: Methods5. Heat Shock ConditionsTransgenic lines were heat shocked for two hours at 33°C on NG plates seeded with 0P50and then allowed to recover at 20°C for 15’. Subsequently, worms were washed off the platesin distilled water, and stained for J3-galactosidase activity as described previously.6. Preparation of RNAUsually four to eight plates loaded with worms was the starting material per RNApreparation. Mixed nematode populations, or embryos prepared by bleaching gravid adults(Emmons et al. 1979), were washed in cold 0.14 M NaCl, and then frozen by dripping theworm suspension into liquid nitrogen. Total RNA was isolated essentially as described byAnton ucci (1985). Frozen worm pellets were powdered with a chilled mortar and pestle andthen dissolved in 1-3 ml of guanidinium solution (7.5 M guanidinium chloride; 25 mM NaCitrate, pH 7.0; 0.1 M f3-mercaptoethanol). The homogenate was passed 2-3 times through asyringe with a 21 gauge needle and then layered over 1 ml of DEPC-treated, sterile separationsolution (5.7 M CsCl; 25 mM sodium citrate, pH 5.0). The RNA was pelleted byultracentrifugation at 42,000 rpm (220,000 x g) for 16 hours. The pellet was thenresuspended in 0.1 to 0.3 ml of sterile, DEPC-treated distilled water, precipitated by additionof 0.1 volumes of sterile, DEPC-treated 3M Na acetate (pH 5.2) and 2.5 volumes of 95 %ethanol. After centrifugation, the RNA pellet was finally resuspended in 50 p.1 of sterile DEPCtreated distilled water and quantitated spectrophotometrically by absorbance at 260 nm.7. Preparation of cDNAFirst strand cDNA was synthesized essentially as described by Graham (Ph.D. thesis,1990) using a modification of procedures outlined by Maniatis et al.(1982) and Frohman et al.(1988). Ten p.g of total RNA suspended in 7.5 p.1 of DEPC treated distilled water was combinedwith an equivalent volume of 40 mM methylmercuric hydroxide and incubated at 20°C for 15’99Chapter Ii: Methodsbefore flash freezing in a dry ice/ ethanol bath. Thirty p.1 of a solution containing 33 p.M DTT,1.66 X MMLV reverse transcriptase buffer (BRL), 1 unit/p.l ribonuclease inhibitor (Promegaor Pharmacia), 833 jiM of each of dATP, dGTP, dCTP, and dTTP, and 2 jig ofoligodeoxyribonucleotide RGO5 or RACE (oligo-dT) was added to the frozen RNA pellet.Subsequently, 4.5 p.1 (900 units) of MMLV reverse transcriptase (BRL) was added to themixture and the reaction was incubated at 37°C for 45’. Excess oligodeoxyribonucleotide wasremoved from the cDNA product by purification on spun filtration columns (MilliporeUltrafree-MC, lOOK). Oligonucleotide RGO5 is 5’ AGGGTTTTCCCAGTCACGAC 3’ and iscomplementary to a sequence in the N-terminal coding region of IacZ.8. Polymerase Chain ReactionsThe cDNA product was amplified by a modification of the procedure of Frohman et al.(1988). Usually one tenth of the eDNA obtained after purification was mixed with 50 pmol ofoligonucleotide OPC6, 1.00 pmol of oligonucleotide SL1, and 45 p.1 of a solution containing 77 j.tMof each of dATP, dGTP, dCTP, and dTTP, 10 mM Tris-HCI pH 8.4, 0.05 % Tween 20, 0.05 %Nonidet P-40, 0.5 mM MgCl, and 1 unit of Taq DNA polymerase (Promega or Pharmacia). TheDNA was amplified on an ERICOMP thermocycler for 35 cycles consisting of 90 sec.denaturation at 94°C, 120 sec. annealing at 59°C, 90 sec. extension at 72°C. The amplified DNAproducts were separated by electrophoresis. on a 2% agarose gel. Oligonucleotide SL1 is a22mer of the sequence 5’ GGTTTAATTACCCAAGTTTGAG 3’, and oligonucleotide OPC6 consists ofthe sequence 5’ GAGGATCCCGATCTCGCCATACAGCGCG3 and is upstream of RGO5.9. Separation and analysis of PCR productsThe amplified DNA products were separated on a 2% agarose /1 X TBE (0.089 M Tris,0.089 M boric acid, 0.002 M EDTA-pH 8.0) gel by electrophoresis at 100 volts in 1 X TBEbuffer.100Chapter II: Methods10. Southern Analysis of PCR ProductsThe PCR products separated on the agarose gel were transferred to a HYBOND-N nylonmembrane according to the procedure of Southern (1975). DNA was immobilized on themembrane by UV crosslinking. MQIF oligonucleotide was labelled with y32PdATP in a reactionusing polynucleotide kinase (Maniatis et a!. 1982) and purified by centrifugation through a G25 fine grade Sephadex spun column. The membrane was hybridized in 10 ml of a solutioncontaining 0.5 % sodium dodecyl sulphate, 5 X Dentiardt’s, 5 X SSC for 10 hours at 50°C,before being washed with the final stringent wash consisting of 0.1 X SSC/0.1 % SDS at 50°C.Oligonucleotide MQIF is a degenerate 32-mer containing a BamHl site in addition to the firstseven codons of ubq-1. Its sequence is 5’ cattggatccgt ATG GANG ATI Tfl/C GTI AANG AC 3’.101Chapter II. ResultsC: RESULTS1. Construction of ubq- 1 -lacZ strainsThe ubq-1-IacZ fusions which were analysed are shown in Figure 14. Ubq938 contains938 bp of sequence upstream of the initial methionine, and extends 108 bp downstream to codon36 of ubq-1 , which is fused to IacZ in the nematode IacZ expression vector pPD16.43 (Fire etal. 1990). Two versions of ubq938, one with an SV4O nuclear localization signal upstream ofIacZ (ubq938NLS) and one without, were used to establish transgenic lines. A deletion seriesextending from the 5’ end of ubq938 was constructed by exonuclease Ill digestion and theresulting products were cloned into pPD1 6.43. None of the deletion constructs possessed theSV4O nuclear localization signal. All IacZ fusions contained 988 bp of ubq-1 3’ non-codingsequence cloned into the 3’ polylinker of pPD1 6.43, to provide the polyadenylation signal andother potentially important elements.Transgenic lines were established by co-injecting the construct of interest, togetherwith a selectable marker, into C. elegans oocytes as described (Fire, 1986). Transformantsselected with pPD1O.41, an unc-22 antisense vector, were identified by their twitchingphenotype (Fire et al., in press) while worms transformed with pRF4, a rol-6 dominantmarker, were identified by their right rolling phenotype (Kramer et. al. 1990; Mello, personalcommunication). The transgenic lines obtained transmitted the marker phenotype at afrequency of 20-80%. Southern analysis (not shown) suggested that the injected DNA formedmixed extrachromosomal arrays, as described by Stinchcomb et al. (1985) and Fire (personalcommunication). Some deletions (Al 83, A523, tX670, A768) were analyzed by transientassays only; in this procedure, the test plasmid alone was micro-injected, and the progeny ofinjected worms were stained. Ubq 938-IacZ, ubqzl827-IacZ, ubqA9O3-IacZ, and ubqPvuIIzl9Q3-IacZ constructs were initially analysed by transient assays, and subsequently inheritable lines.102Chapter II: Results5SSSt.1SELECTION PLASMIDHSE? GIC HSE? GATA 4, NLSiacz ubq 3 pRF4 pPDlO.41 NONEUbq938NLS I liii - - LI —Ubq938 I — + +U bq LS. 1 8 3 I tEJ — — +U bq 5 23 I __IEJ — — +U bq 670 I —_JEJ — — +Ubq Z.768 I — — +Ubq.827____— + +UbqL903— + +L [E] — + +Ubq-Pvu II 903Fig. 14. Construction of ubq-1-IacZ fusions (not drawn to scale). Ubq938-NLSIacZ contains1044 bp of ubq- 1 sequences 5’ sequences extending from a Sal I site 938 bp upstream of theinitial methionine to an EcoRl site 103 bp downstream of the methionine blunted into the BarnHI site of the lacZ expression vector, .pPD16.43 (Fire et aL, 1990). To construct ubq938-lacZ, a Hind Ill/Kpn I fragment containing the Sal l/Eco RI ubq-1 fragment of ubq938NLS-IacZwas directionally cloned into pPD16.43 which had been digested with Hind Ill and Kpn I toremove the SV4O nuclear localization signal. Thus ubq938-lacZ and all of the deletions are in-frame translational fusions containing th first 36 amino acids of ubiquitin but are devoid of theSV4O NLS. Deletions extending from the Sal I site of ubq938-lacZ were generated byexonuclease Ill digestion. See Methods for details. The horizontal arrow indicates the directionof transcription from the putative start site based on Si analysis (5’SS). The vertical arrowindicates the 3’ acceptor site for trans-splicing with SL1 RNA which is located 6 bp upstream ofthe initial methionine of ubq-1. All of the deletions possessed the 3’ acceptor site. HSE,putative heat shock element; GIG, sequence of cytosine residues.; GATA, TATA-like sequence,GAATAA. Selection schemes tested with each lacZ fusion are designated by plus signs at the rightof each fusion. Two plasrnids were used to select for transformed lines, pRF4, a rol-6 marker,and pPD1 0.41, an unc-22 vector. In transient assays the test plasmid was injected alone.103Chapter II. Results2. Expression of the ubq938-lacZ transgene is constitutive but shows developmental tissuespecificityInitially transgenic lines were established which carried ubq938 without the SV4Onuclear localization signal. The expression of the ubq938-lacZ transgene was constitutive at allstages of development in these strains. While some of the f-galactosidase staining was localizedin the nucleus, the cytoplasm also stained extensively, making cell identification difficult. Thisproblem was resolved by inclusion of the SV4O nuclear localization signal (Figure 15).Interestingly, there were differences in both the distribution and intensity of 3-galactosidasestaining between embryonic and post-embryonic stages. Embryos, from gastrulation onward,stained intensely within 15-30 minutes of the start of the incubation. In contrast, overnightincubation in the staining solution was usually required to achieve detectable somaticexpression in post-embryonic stages. This suggests that embryos carrying the ubq938-lacZtransgene constitutively produce a greater quantity of f3-galactosidase relative to somatictissues. Moreover, the distribution of -galactosidase activity was more limited in the post-embryonic stages. For example, in one strain, P036 (ubq938NLS-lacZ/pRF4), one fifth ofstaining Li larvae efficiently expressed the transgene in body muscle, hypodermis, pharyngealtissue, and nervous tissue, while expression in adults was usually limited to body muscle(Figure 15). Approximately 20% of all F2 (derived from a single pre-selected rollinghermaphrodite) embryos, Li and 12 larvae were stained, while the frequency of somatic tissueexpression was a mere 5% in adults. However, 16% of P036 adults carried stained embryos.Subjecting ubq938-lacZ strains to a two hour heat shock treatment at 33°C had nodetectable effect on the intensity or distribution of staining at any stage in development.104Chapter II: ResultsFig. 15. In situ staining of j3-galactosidase activity in P036, a transgenic strain carrying anextra-chromosomal array composed of ubq938NLS and pRF4. (a) Expression in an adult afterovernight incubation in stain. Gastrulation and later staged embryos are intensely stained whileonly a few body muscle nuclei stain in the somatic tissues. Magnification 150X. (b) Staining incomma stage (lower right) and pretzel stage (upper left) embryos. Magnification 370X. (C)Staining of primarily body muscle nuclei in an Li larva. A few hypodermal nuclei are alsostained. Magnification 370X.105106b.•,.‘,Chapter II: Results3. Expression is not diminished until 827 bp of sequence has been deleted from ubq938-lacZProgressively larger deletions extending from the 5’ end of ubq938-lacZ were tested bytransient assays to determine if they altered the pattern of expression. Expression of theubqzll83-lacZ and ubqzi523-lacZ transgenes was intense in embryos, while somatic expressionin post embryonic stages was limited often to the pharynx or the body muscle in the head.Transient assays of ubqA67O-lacZ and ubqA768-lacZ also failed to reveal any significantreduction in expression (Figure 16). Surprisingly, expression was not significantly altereduntil 827 bp of sequence had been removed to a point 330 bp downstream from thetranscriptional start, or about 120 bp upstream of the initial methionine (Figure 16). Incontrast to ubq93S-lacZ embryos, which were usually stained to saturation within one hour(Figure 1 7a), ubq4827-lacZ embryos required four to six hours to reach comparable levels ofstaining (Figure 17e). Further removal of sequences to a point 30 bp upstream of the 3’acceptor site for SL1 (ubqA9O3-lacZ) drastically reduced expression (Figure 17g,hJ).Embryos carrying the ubqi903-IacZ transgene stained marginally after overnight incubationand only the occasional Li larva stained (lightly) in the pharynx (Figure 17i). No somaticexpression was observed in later larval stages.These results suggest that sequences between the ubq1827-IacZ and the ubqzl9O3-lacZbreakpoints contribute to expression while sequences upstream of the transcriptional start site(including a GIG rich block and a GATA box) are not required for embryonic expression of ubq1. To test this hypothesis I assayed animals carrying ubq-pvulli.1903-lacZ, a construct with thefirst 538 bp of ubq938-lacZ fused to ubq z1903-lacZ. Embryos carrying ubq-pvulIzl9O3-lacZshowed expression similar to that of ubqA9O3-lacZ (Figure 17j,k,l). To further analyze theubq4827-lacZ, ubqzl9O3-lacZ, and ubq-pvu114903-lacZ fusions, transgenic strains carryingthese fusions were generated using the unc-22 antisense vector, pPDi 0.41. The patterns ofexpression observed for these strains resembled those obtained in the transient assays.107Chapter II: Results5SSSL1IacZ EXPRESSIONHSE’? G/C HSE? GATA ATG IacZ ubq 3 1 hr. 6 hr. 16 hr.IJbq 938_iiI + + +U bq . 1 8 3___________________________________________i:ij +U bq 5 23 I +Ubq. 670 I__.IiI. +Ubq768i_______ JEi +Ubq827 I + + +Ubq.903—c: — — +____EE1—— +Ubq-Pvu II 903Fig. 16. Deletion analysis of the ubq938-lacZ transgene. Presence of j3-galactosidase activity intransgenic strains and/or in transient assays after varying durations of incubation in stain isindicated by plus signs to the right of each construct.108Chapter 11: ResultsFig. 17. Expression of ubq-1-IacZ deletion transgeries. Transgenic animals were incubated forone, six, or sixteen hours in a solution containing Xgal (Fire et al. 1990) before beingpermanently mounted in 80% glycerol, 20 mM Tris-pH 8.0, 200 mM sodium azide forexamination. (a,b,c): Staining of PC18, a ubq 938-lacZ/pPD1O.41 strain. (a) After one hour,embryos within adults are saturated with stain. Staining of somatic tissues in larvae (upperleft of picture) has also begun. (b) After six hours, 3-galactosidase activity is increasinglyapparent in the body muscle of larvae (bottom picture). (C) After 16 hours, expression inbody muscle is obvious in an L3 larva, (top) and in the head of an adult (bottom). In addition,many Li larvae are saturated with stain (centre). (d,e,f): Staining of PC4, a ubq827-lacZ/pPDi 0.41 strain. (d) After one hour, individual nuclei in pretzel stage embryos are staining.(e) At six hours, staining in newly hatched Li larvae is obvious (top) and embryos aresaturated (bottom). (f) 16 hours. Somatic tissue expression in the pharynx of an L4 (bottomleft); throughout body of Li (top left); and body muscle in posterior of Li (top right).Saturated embryo (bottom right)(g,h,i): Staining of PCi, a ubq903-lacZIpPD10.41 strain.(g,h) No expression is observed in either somatic tissues or embryos even after six hours. (i)After 16 hours incubation, weak expression is visible in the terminal bulb of the pharynx of anL2 (top) and in a pretzel embryo. (j,k,l): Staining of P017, a ubq-PvullA9O3-IacZ/pPDiO.4i strain. (j,k) Even after six hours no expression is visible. (I) After 16hours individual nuclei in pretzel stage embryos (bottom) are staining and occasionallymoderate staining (top ) is observed.109OTTpiH9l.NHDaNH9I.L8vbqn9E6bqn‘ItI1H91.9IHLIIi,IH9Rco6vbcln eo6VIInAd—bqnChapter II: Results4. The ubq938-lacZ, and ubq4827-lacZ transcripts are trans-splicedTo determine if the ubq-1-lacZ fusion mRNAs, like those of ubq-1, were trans-spliced,the polymerase chain reaction (PCR) was utilized to selectively amplify only trans-spliced lacZtranscripts (Figure 18). Initially, the validity of this approach was verified by controlexperiments. First strand cDNA from both ubq-1-lacZ and endogenous ubq-1 messages wasgenerated from total RNA using the lacZ -specific oligonucleotide primer RGO5, or the oligo-dTprimer RACE and reverse transcriptase. After the removal of excess primer, the cDNA wasintroduced into a polymerase chain reaction with various combinations ofoligodeoxyribonucleotides and the resulting products separated by gel electrophoresis (Figure19). When MQIF, a degenerate oligonucleotide encoding the first seven amino acids of ubiquitin,was combined with the lacZ oligo, OPC6, distinct bands of the expected sizes were observed inPC36 and PC4 experiments ( Figure 19, Lanes 2 and 3) while the wild type negative controlexperiment revealed faint background bands (Lanel). Controls in which either SL1 or OPC6alone was added to the reaction occasionally yielded very faint background bands (Figure 19,lanes 4 to 9), indicating that the majority of RGO5 or RACE had been removed.To amplify only trans-spliced ubq-1-lacZ mRNAs, first strand cDNA was amplifiedwith oligonucleotides specific for spliced leader 1 (SL1) and IacZ (OPC6, art oligonucleotideinternal to RGO5). The products of these reactions were separated on a 2% agarose gel and areshown in Figure 20 a. Bands of the appropriate sizes were consistently observed for ubq938-IacZ and ubq827-IacZ; however, faint non-specific bands were somelimes observed when wildtype cDNA was included as a control. To verify the identity of the bands observed in the ubq938-IacZ and ubqx827-IacZ experiments, Southern analysis was carried out using probeMOlE, (Figure 20b). MOlE hybridized only to the putative ubq938-lacZ and ubqz\827-lacZbands, suggesting that they represent amplified trans-spliced lacZ transcripts. The results forubq9O3-IacZ and ubq pvuIlA9O3-lacZ were ambiguous: usually no amplified bands wereobserved (Figure 20), but occasionally faint bands of varying sizes were seen.112Chapter II: Results-- 4-SLt MUll RACE ubq-1IHAA- —*4-4- 4-Sil MUll OPC6 Rl85 AlICE ubq-I/lacZ“AnnARNARGBSReverse TranscriptasedilNilOPC6I sil4 TaqRun products on 2% gelblat. probe with MUll.Fig. 18. Amplification of trans-spliced ubq-1-IacZ RNA. First strand cDNA was reversetranscribed from total nematode RNA with either oligodeoxyribonucleotide RGO5 or RACE (oligodT). The cDNA was introduced into a polymerase chain reaction with oligodeoxyriboucleotidesOPC6, and SL1 and Taq DNA polymerase. The PCR products were separated by electrophoresis,transferred to a nylon membrane and probed with MQIF, an oligodeoxyribonucleotidecorresponding to the first seven codons of ubiquitin.113Chapter II: Results910 bp540426,409235166Fig. 19. Testing the integrity of the PCR amplification scheme. Various cDNA5 were tested inpositive and negative control experiments, and analyzed by electrophoresis on a 2 % agarose gel.See text for details. Lanes: 1, Wild type (N2) cDNAI amplified with MQIF and OPC6; 2, PC36(ubq938NLS-lacZ)/ MQIF and OPC6; 3, PC4 (ubq827-IacZ)/ MQIF and OPC6; 4, Wild type/SL1; 5, PC36 (ubq938NLS-lacZ)/ SL1; 6, PC4 (ubqA827-lacZ)/ SL1; 7, Wild type! OPC6;8, PC36/ OPC6; 9, PC4/ OPC6; 10, Blank; 11, pUC13 digested with Ddel molecular weightmarker. The arrows indicate bands of the appropriate size present in the positive controls(lanes 2 and 3). The amplified product of PC36/ MQIF and OPC6 was expected to be 263 bpwhile the PC4 product was expected to be only 218 bp since it does not possess the SV4O NLS.1 234578 91011114*412345678Fig.20. Analysis of PCR products amplified from ubq-1-IacZ RNA. RNA was reversetranscribed with MMLV reverse transcriptase and RGO5. First strand cDNA was amplified withTaq Polymerase and the oligonucleotides, SL1 and OPC6. (a) The amplified products wereseparated by electrophoresis on a 2% gel. Lanes: 1, Molecular weight marker generated bydigestion of plasmid pUC13 with restriction endonuclease Ddel; 2, Wild type control (N2); 3,Wild type (N2) control (In this experiment first strand cDNA was primed with oligo dT(RACE)); 4, PC36 (ubq938NLS-lacZ); 5, PC4 (ubqz827-IacZ); 6, PCi (ubqz9O3-lacZ); 7,PC2 (ubq9O3-lacZ); 8, PC17 (ubq-PvullA9O3-IacZ). The arrows indicate bands of theappropriate size present in lanes 4 and 5. The amplification product of the PC36 (ubq938NLS)experiment was expected to be 278 bp whereas the PC4 (ubqA827) PCR product was expectedto be 233 bp since the latter construct does not possess the SV4O NLS. (b) Southern blot of thegel in (a) probed with oligo MQIF. Lanes: 1 to 8 as in (a).115Chapter II: Results45678ab-4910540426,409Chapter 11: DiscussionD. DISCUSSION1. Expression of ubq-1-IacZ transgenes in nematodes.The ubq938-lacZ fusion was expressed in a tissue-general manner in embryos and in20% of newly hatched transgenic Li larvae. Surprisingly, the transgene was not expressedextensively in the somatic tissues of later stages. Given the known functions of ubiquitin, andthe fact that ubq-1 seems to be the only polyubiquitin gene in C. elegans, the expectation wasthat it would be expressed in most adult tissues. Indeed, using P element mediatedtransformation of lacZ fusions, Lee et al. (1988) found that the polyubiquitin gene ofDrosophila is expressed in all tissues throughout development. In Dictyostelium and chicken,polyubiquitin genes are developmentally regulated, but each of these organisms possesses morethan one polyubiquitin gene (Giorda and Ennis, 1987; Rocamora and Agell, 1990).Furthermore, Northern analysis of the endogenous ubq-1 gene of C. elegans at various stages ofdevelopment previously revealed that levels of ubq-1 RNA remain relatively constantthroughout development (Graham et al. 1989).Thus, the tissue specificity in post-embryonic stages is surprising. The hypodermalexpression observed occasionally in PC36 (ubq938-lacZ/pRF4) larvae (see Figure 15) maybe an artefact due to enhancer sequences in the rol-6 vector (pRF4). Such expression has notbeen confidently documented in ubq938-lacZ/pPD1O.41 transgenic strains. On the other hand,the observed muscle expression seems to be real since animals carrying ubq938-lacZconstructs expressed the transgene in these nuclei regardless of which selection, if any, wasemployed.It is conceivable that ubq-1 sequences necessary for somatic expression in later stageswere not included in the ubq938-lacZ construct. I therefore tested a construct which containeda further 763 bp of upstream sequence (Ubql7Ol-IacZ) in transient assays and found thatexpression of this transgene was similar to that of ubq938-lacZ. In addition to the 5’sequences, all of the ubq-1-lacZ fusions used in this study contained 988 bp of ubq-1 3’ non116Chapter II: Discussioncoding sequence which included the eridogenous polyadenylation signal. Thus it seems unlikelythat critical 5’ or 3’ regulatory sequences were excluded from these constructs.The endogenous ubq-1 gene contains four typical cis-spliced introns in the first,fourth, seventh, and tenth ubiquitin coding repeats. None of these introri sequences werepresent in any of the IacZ fusions. It is possible that one or more of the ubq-1 introns containsenhancer sequences which are necessary for maintaining expression in the post embryonicstages of C. elegans. Alternatively, the normal position of ubq-1 in the genome relative to othersequences may be important for optimal expression.Of the many functions attributed to ubiquitin, its role in targetting proteins fordegradation is the best characterized. This raises the possibility that endogenous ubiquitinmay have recognized the fusion protein as being abnormal ubiquitin and targetted it forproteolysis by the ATP dependent non-lysosomal proteolytic pathway. Using ubiquitin-lacZfusions, Bachmair et al. (1986,1989) showed that ubiquitin--gaIactosidase fusion proteinswere rapidly de-ubiquitinylated in vitro to release free ubiquitin and functional 13-galactosidase, and that the half life of the released f3-galactosidase depended upon the identity ofthe amino acid present at the mature amino terminus of the enzyme (the “N-end rule”). Theubiquitin moiety was almost always cleaved off the fusion protein after its carboxyl terminalglycine no matter which amino acid residue was at the amino end of f3-galactosidase. Only whena proline residue was present at the amino terminus of f3-galactosidase was the fusion proteinnot de-ubiquitinylated. In this case, the fusion protein was rapidly degraded.I propose that the fusion proteins produced in this experiment could not be deubiquitinylated due to lack of the ubiquitin carboxyl terminus and that these proteins may havebeen recognized as aberrant and degraded by the ATP-dependent proteolytic pathway.Transformations of constructs which contain either one complete ubiquitin coding sequencefused to IacZ, or no ubiquitin coding sequences at all, should yield more extensive expression ifthis hypothesis is correct. Intriguingly, the ubq-IacZ fusion which Lee et al. (1988) used to117Chapter II: Discussiontransform Drosophila contained no ubiquitin coding sequences but all of the upstreamuntranslated sequences.2. Ubq- 1 expression is not significantly heat inducibleExpression of the ubq938-lacZ fusion was neither more extensively distributed norincreased in level after a two hour heat shock treatment, in agreement with previous results ofGraham et al. (1989) based on Northern analysis of ubq-1 transcripts. In contrast,polyubiquitin genes of several organisms including chicken (Bond and Schlesinger 1985), yeast(Saccharomyces cerevisiae, Tanaka et al. 1988) and Drosophila (Lee et al. 1988) are inducedupon heat shock. In Drosophila, however, the induction was weak and variable from experimentto experiment. Since modest increases in expression would be difficult to detect by eitherNorthern analysis or in situ iacZ staining, we cannot rule out the possibility that ubq- 1 may beweakly heat inducible There exist two regions upstream of ubq-1 which possess sequencesimilarity to heat shock elements. The more distal of these, at -827 (Graham et al. 1989),contains two nGAAn motifs. It has been shown that while only two nGAAn motifs are required tostably bind heat shock factor in vitro (Xiao and Lis, 1988; Perisic et al., 1989), three suchmotifs are necessary to produce a strongly inducible heat shock element in vivo (Perisic et al.,1989; Xiao et al., 1991). By this criterion this sequence at best may represent a weak heatshock element; furthermore, it lies approximately 370 bp away from the putativetranscription start site at -455. The proximal sequence at -650 contains only one nGAAn motifand thus is an unlikely candidate for a functional heat shock element. It is thus not surprisingthat ubq-1 expression is not conspicuously induced by heat shock.This does not negate the possibility that the ubiquitin mediated proteolytic pathway isheat inducible in C. elegans. Increased demand on this pathway after heat stress could be met byincreasing the rate of ubiquitin conjugation to substrates and by increasing the quantity ofavailable monomeric ubiquitin by post translational processing of polyubiquitin. Theobservations that two genes encoding ubiquitin conjugating enzymes in yeast are heat inducible118Chapter II: Discussion(Seufert and Jentsch, 1990) and that the ubq-1 gene of C. elegans encodes a polyprotein of 11ubiquitin molecules (Graham et aL, 1989) provides credence to this idea.3. Trans-splicing of ubq- 1-lacZ transcriptsThe results of the PCR analysis suggest that transcripts from the ubq938-lacZ andubqz827-IacZ transgenes, like those of ubq-1, are correctly trans-spliced. No evidence fortrans-splicing of ubqt903-lacZ or ubq PvullA9O3-lacZ was obtained, suggesting that thenecessary signals had been removed. This raised the possibility that the reduction in expressionobserved with these constructs resulted from a failure in trans-splicing. Recent evidencehowever suggests that trans splicing by SL1 occurs whenever a 3’ acceptor site is presentunaccompanied by a 5’ donor site (Conrad et al., 1991). The only apparent requirement is foran intact 3’ splice acceptor site. Since both ubq903-lacZ and the larger fusionubqPvuIl903-lacZ retained the 3’ splice site and showed equivalent levels of expression, it isunlikely that the reduced expression seen in these transgenes is the result of a failure in transsplicing. While PCR is a sensitive technique, presumably capable of amplifying singlemolecules, it is possible that there simply was not enough starting material in ubq903-lacZand ubqPvull9O3-lacZ animals to allow the detection of trans-spliced products, especiallygiven the initial reverse transcriptase step involved.4. Analysis of the ubq-1 promoter.A more likely explanation for the observed reduction in expression from the ubqA9O3-lacZ and ubq-PvulIii9O3-lacZ constructs is the loss of elements of the ubq-1 promoter. Thebreakpoint of ubq4903-lacZ is only 36 bp upstream of the initial methionine, and more than400 bp downstream from a GAATAA sequence and putative transcriptional start site. While thelimited expression in ubqA9O3-IacZ and ubqPvuIli9O3-lacZ transgenics suggests that thepromoter has not been completely destroyed, it seems likely that some regulatory sequenceshave been lost, and that these sequences are contained within the region between the ubqii827-119Chapter II: DiscussionIacZ and ubq4903-lacZ breakpoints. This region lacks any obvious sequence similarity toknown promoters or enhancers.Additionally, it is conceivable that the ubq-1 promoter has been completely destroyedand that unc-22 enhancer sequences present in mixed arrays of pPD1O.41 with ubqzl9O3-IacZor ubq-Pvu114903-IacZ promotes the spurious expression of IacZ sequences observed in thesestrains.Sequences upstream of the transcriptional start site, including the TATA-like sequence(GAATAA) and a long stretch of cytosine residues, appear to be unnecessary for embryonicexpression of ubq-1 . A similar situation has been documented with the C. elegans unc-54 gene;in that case a tissue specific enhancer is sufficient for proper expression (A. Fire and S. White-Harrison, personal communication).The constructs ubqA523 : :lacZ, ubq67O ::lacZ, ubqA768 : :lacZ, ubqA827::lacZ andubq\9O3::lacZ lack the wild type transcriptional start site and presumably utilize cryptic startsites present within the remaining 5’ untranslated sequences or possibly within the vector.While it would be desirable to determine the actual start sites of these transgeneS, thisexperiment is difficult for several reasons. Firstly, the fact that ubq-1::IacZ transcripts arerapidly trans-spliced means that very little of the total ubq-1 RNA is present as unprocessedtranscripts. Secondly, incomplete transmission of the transgenes as extrachromosomal arrayscomplicates isolation of the mass quantities of transgenic animals necessary for RNA analysis.Finally, expression of f3-galactosidase. is so limited in both ubq9O3::lacZ, which does notpossess the endogenous transcription start site, and in ubqPvull9O3::lacZ, which does containthe ubq-1 transcriptional start site, that the quantity of available material for RNA analysiswould be reduced even further.120Chapter 11: Discussion5. Conclusions(1) The ubq938-lacZ transgene was expressed constitutively in embryos but showeddevelopmental tissue specificity in the post embryonic stages. Somatic expression in L2, L3,L4 larvae and adults was usually confined to body or pharyngeal muscle.(2) Heat shock did not alter the intensity or distribution of f-galactosidase staining inubq938-iacZ transgenic animals. This supports previous evidence (Graham et al., 1989) thatubq- 1 is not heat inducible at the transcriptional level.(3) Deletion analysis of the ubq938-IacZ transgene suggested that the region betweenthe ubqzl76S-lacZ and ubqA9O3-lacZ breakpoints contains upstream activating sequences. Adownshift in expression was first observed in ubqii827-lacZ transgenic animals and a furtherreduction was observed in ubqzl9O3-IacZ animals.(4) PCR analysis suggests that ubq938-IacZ and ubqzl827-IacZ transcripts areefficiently trans-spliced by SL1 in vivo, and that there are no specific signals for transsplicing by SL1 upstream of the ubqzl827 breakpoint.6. Future ProspectsThe results of this study suggested that sequences contained between the ubq1827-iacZand ubqzi9O3-IacZ breakpoints are important elements in the promotion of ubq-1 expression.Future experiments should include a more detailed analysis of this region to isolate thesequences constituting the ubq-1 promoter. In addition the potential role of introns in geneexpression could be investigated by testing larger fusions which include entire ubiquitinrepeats inclusive of introns.I had anticipated that ubq-1 would possess a strong tissue general constitutive promoterwhich might be confined to a small region, and this clearly was not the case. However, if futuretransformation experiments determine that sequences within the introns contain tissue specificenhancers, it may be possible to ligate all of these elements together to form a relatively121Chapter II: Discussioncompact, tissue general promoter which could be used to drive constitutive expression of anycoding region in C. elegans.Currently, the ubiquitin-ribosomal fusion genes are being cloned and characterized byDon Jones in this laboratory. It would be interesting to determine the spatial and temporalexpression pattern of these genes in C. elegans using the methods described in this thesis forcomparison with that of ubq-1. In addition, deletion of the ubq-1 gene by a recently describedPCR mutagenesis procedure (Wood, C. elegans meeting abstracts, 1991) or over-expression ofubq-1 under control of a heat shock promoter may assist in defining its importance forproviding ubiquitin monomer in cells under various conditions.Finally, mutagenesis of ubq-1 and re-introduction into the nematode genome bytransformation may define sites essential for the proper folding and function of ubiquitin.122ReferencesIll: REFERENCESAlbertson, D.G. and J.N. Thomson. 1976. Phil. Trans. R. Soc. Lond. B. ZZ:287-297.Amin, J., J. Ananthan and R. Voellmy. 1988. Mol. Cell Biol. j 3761 -3769.Ananthan, J., A.L. Goldberg and R. Voellmy. 1986. Science 2:522-524.Antonucci, T.K. 1985. Recombinant DNA Techniques fi:22-24. Univ. of Michigan.Arrigo, A.-P., J.P.Suhan and W.J. Welch.1988. Mol. Cell Biol. fi:5059-5071.Ashburner, M. and J.J. Bonner. 1979. Cell ui 241-254.Bachmair, A. and A. Varshavsky. 1989. Cell :1 019-1032.Bachmair, A., D. Finley, and A. Varshavsky. 1986. Science. 4:179-186.Baker, R.T., and P.G. Board. 1987. Nucleic Acids Res. 1:443-463.Ball, E., C.C. Karlik, C.J. Beall, D.L.bSaville, J.C. Sparrow, B. Bullard, and E.A. Fryberg. 1987.Cell L:221 -228.Banerji, S.S., K.Laing and R.l. Morimoto. 1987. Genes 0ev. 1:946-953.Bardwell, J.C. and E.A. Craig. 1984. Proc. NatI. Acad. Sci. USA 1:848-52.Bartel, B., I. Wunning, and A. Varshavsky. 1990. EMBO J. .:3179-3189.Baxter , R. C. 1985. Endocrinology .jji:650-655.Beaulieu, J.-F., A.-P. Arrigo and R.M. Tanguay. 1989. J.Cell Sci. :29-36.Beckmann, R.P., L.A. Mizzen, and W.J. Welch. 1990. Science 4.a:850-854.Behlke, J., G. Lutsch,.M. Gaestel and H. Bielka. 1991. FEBS Lett. :119-122.Berger, E.M. and M.P. Woodward. 1983. Exp. Cell Res. .141:437-442.Bienz, M. 1984. Proc. NatI. Acad. Sci. USA i:31 38-3142.Bienz, M. and J.B. Gurdon. 1982. Cell :811-819.Birnboim, H.C. and J. Doly. 1979. Nucleic Acids Res. Z:1 51 3-1 523.Blumenthal, T. and J. Thomas. 1988. Trends Genet. 4:305-308.Bond, U. 1988. EMBO J. Z:3509-3518.Bond, U. and M.J. Schlesinger. 1985. Mol. Cell. Biol. .:949-956.123ReferencesBond, U. and M.J. Schlesinger. 1986. Mol. Cell. Biol. .:4602-461 0.Bond, U. and M.J. Schlesinger. 1988. Adv. Genet. 24:1-29.Bonner, J.J. C. Parks, J. Parker-Thornberg, M.A. Mortin and H.R.B. Pelham. 1984. Cell3.1:979-991.Brenner, S. 1974. Genetics ZZ:71-94.Browder, L. W., M. Pollock, J.J. Heikkila, J. Wilkes, T. Wang, P. Krone, N. Ovsenek, and M.KIoc. 1987. Dev. Biol. 1a4.:191-199.Brown, l.R., D.G. Low and L.A. Moran. 1985. Neurochem. Res. J&:1 277-1284.Bruzik, J.P., K. Van Doren, D. Hirsh, and J. Steitz. 1988. Nature 3.3.:559-562.Bukau, B. and G. Walker. 1989. J. Bacteriol. jii:6030-6038.Busch, H. 1984. Methods Enzymol. iQfi:238-262.Candido, E.P.M., D. Jones, O.K. Dixon, R.W. Graham, R.H. Russnak and R.J. Kay. 1989. Genome31:690-697.Chandrasekhar, G.N., K.Tilly, C.Woolford, R. Hendrix and C. Georgopoulos. 1986. J. Biol. Chem.ZIt:1 2414-12419.Chappell, T.G., W.J. Welch, D.M. Schlossman, K.B. Palter, M.J. Schlesinger, and J.E. Rothman.1986. Cell 4..:3-13.Chau, V., J.W. Tobias, A. Bachmair, D. Marriott, D.J. Ecker, O.K. Gonda, and A. Varshavsky.1989. Science 243.. :1576-1583.Chirico, W. J., M.G. Waters and’ G. BlobeLl988. Nature 3.2.: 805-810.Chrétien, P. and J. Landry. 1988. J. Cell. Phys. J.aZ:157-166.Christiansen, E.N. and E. Kvamme. 1969. Acta. Physiol. Scand. Z:472-484.Ciechanover, A., D. Finley, and A. Varshavsky. 1984. J. Cell Biochem. 24:27-53.Ciechanover, A., J.A. DiGiuseppe, B. Bercovich, A. Orian, J. D. Richter, A.L. Schwartz, and G. M.Brodeur. 1991. Proc. NatI. Acad. Sci. USA .a:139-143.Ciechanover, A., Y.Hod, and A. Hershko. 1978. Biochem. Biophys.Res.Commun. .i:1 100-1104.Cbs, J., T. Westwood, P.B. Becker, S. Wilson, K. Lambert and C. Wu.1990. Cell 3:1085-1097.Cohen, R. S. and M. Meselson. 1985. Cell 4.3:737-746.124ReferencesCollier, N. C., J. Heuser, M.A. Levy , and M.A. Schlesinger.1988. J. Cell Biol. 1Q.:1131-11 39.Conrad, R., J. Thomas, J. Spieth, and T. Blumenthal. 1991. Mol. Cell. Biol. 11:1921-1926.Cooper, P. and T.-H. D. Ho. 1987. Plant Physiol. 4j 1197-1203.Copeland, C. S., R.W. Doms, E.M. Bolzau, R.G. Webster, and A. Helenius. 1986. J. Cell Biol.103: 1179-1191.Corces, V., R. Holmgren, R. Freund, R. Morimoto and M. Meselson. 1980. Proc. NatI. Acad. Sci.USA ZZ:5390-5393.Costlow, N. and J.T. Lis. 1984. Mol. Cell. Biol. :1853-1863.Coulson, A., J.E. Sulsion, S. Brenner and J. Karn. 1986. Proc. Nati. Acad. Sci. USA :7821-7825.Coulson, A., R. Waterston, J. Kiff, J. Sulston and V. Kohara. 1988. Nature .5:184-186.Davie, J.R. and L.C. Murphy. 1990. Biochem. 2.: 4752-4757.DeMarzo, A. M., C..A. Beck, S.A. Onati, and D.P. Edwards. 1991. Proc. NatI. Acad. Sd. USA :72-76.Deshaies, R. J., B.D. Koch, M. Wemer-Washburne, E.A. Craig, and R. Sheckman. 1988. Naturea2.: 800-805.Dickson, J.A. and B.E. Oswald. 1976. Br. J. Cancçr. 34:262-271.DiDomenico, B.J., G.E. Bugiasky and S. Lindquist. 1 982a. Cell 34:593-603.DiDomenico, B.J., G.E. Bugiasky, and S. Lindquist. 1982b. Proc. Nati. Acad. Sd. USA Z:6181-6185.DiNocera, P.P. and l.B. Dawid. 1983. Proc. Nail. Acad. Sci. USA :7095-7098.Dixon, D. K., D. Jones, and E.P.M. Candido. 1990. DNA Cell Biol. a:1 77-191.Dohmen, R.J., K. Madura, B. Bartel and A. Varshavsky. 1991. Proc. Nail. Acad. Sci. USA.:7351 -7355.Dudler, R. and A. A. Travers. 1984. Cell 3..:391 -398.Dworkin-Rastl, E., A. Shrutkowski, and M. B. Dworkin. 1984. Cell 3.a: 321-325.Ellis, H. M. and H.R. Horvitz. 1986. Cell 44: 817-829.Emmons, S. W., M.R. Klass, and 0. Hirsh. 1979. Proc. Natl. Acad. Sci. USA Z:1333-1337.Emmons, S.W., B. Rosenzweig and D. Hirsh. 1980. J. Mol. Biol. 144:481-500.125ReferencesEpstein, H. F., R.H. Waterston, and S. Brenner. 1974. J. Mol. Biol. Q:291-300.Falkner, F.-G., H. Saumweber and H. Biessman. 1981. J. Cell Biol. j..:1 75-1 83.Findly, R.C. and T. Pederson. 1981. J. Cell Biol. :323-328.Findly, R.C., R.J. Gillies and R.G. Schulman. 1983. Science. 2.11:1223-1225.Finley, D., A. Ciechanover, and A. Varshavsky. 1984. Cell a.Z:43-55.Finley, 0., B. Bartel, and A. Varshavsky. 1989. Nature :394-400.Finley, D., E. Ozkaynak, and A. Varshavsky. 1987. Cell 4a: 1035-1046.Fire, A. 1986. EMBO J. .:2673-2680.Fire, A., S. White Harrison, D. Albertson, and 0. Moerman. 1991. Development in press.Fire, A., S. White Harrison, and D. Dixon. 1990. Gene 3.: 189-198.Fried, V.A., H.T. Smith, E. Hildebrandt, and K. Weiner. 1987. Proc. NatI. Acad. Sci. USA4..:3685-3 689.Frohman, M.A., M.K. Dush, and G.R. Martin. 1988. Proc. NatI. Acad. Sci. USA :8998-9002.Gaestel, M., W. Schroder, R. Benndorf, C. Lippmann, K. Buchner, F. Hucho, V.A. Erdmann, and H.Bielka. 1991. J. Biol. Chem. 2fi.:14721-14724.Gaitanaris, G.A., A.G. Papavassiliou, P. Rubock, S.J. Silverstein and M.E. Gottesman. 1990. CellLL:1 013-1 020.Gaterman, K.B., G.H. Rosenberg and N.F. Kaufer. 1988. Biotechniques fi:951-952.Gething, M.-J., K. McCammon, and J. Sambrook. 1986. Cell 4..:939-950.Gilmour, D.S. and J.T. Lis. 1986. Mol. Cell. Biol. fi:3984-3989.Giorda, R. and H.L. Ennis. 1987. Mol. Cell Biol. .:2097-2103.Glaser, R.L., M.F. Woitner and J.T. Lis. 1986. EMBO J.:747-754.Glotzer, M. A.W. Murray, and M.W. Kirschner. 1991. Nature a.4..:132-138.Goebl, M.G., J. Yochem, S. Jentsch, J.P. MaGrath, A. Varshavsky and B. Byers. 1988. ScienceZJ...: 1 331 -1 335.Goldenberg, C.J., Y. Luo, M. Fenna, R. Baler, R. Weinmann and R. Voellmy. 1988. J. Biol. Chem.aa:1 9734-1 9739.Goldschmidt, R. 1935. Abst. u. Vererb. 1:38-69; 70-131.126ReferencesGoldstein, G., M. Scheid, U. Hammerling, E.A. Boyse, D.H. Schlesinger, and H.D. Niall. 1975.Proc. Natl. Acad. Sd. USA ZZ:1 1-15.Goloubinoff, P., A.A. Gatenby and G.H. Lorimer. 1989. Nature. Z:44-47.Gorin, E. and H.M. Goodman. 1985. Endocrinology JJ:1 796-1 805.Grabowski, P.J., R.A. Padgett and P.A. Sharp. 1984. Cell aZ:41 5-427.Graham, R.W., D.Jones, and E.P.M. Candido. 1989. Mol. Cell. Biol. :268-277.Graham, R.W., K. Van Doren, S. Bektesh, and E.P.M. Candido. 1988. J. Biol. Chem.Z3: 1 041 5-1 041 9.Grossman, A.D., D.B. Straus, W.A. Walter and C.A. Gross. 1987. Genes Dev. 1:1 79-1 84.Haas, A.L., P. Ahrens, P.M. Bright, and H. Ankel. 1987. J. Biol. Chem. ZZ: 11315-11323.Hammond, G.L., Y.-K. Lai and C.L. Markert. 1982. Proc. Natl. Acad. Sci. USA Z:3485-3488.Hanahan, D. 1983. J. Mol. Biol. j.a: 557-580.Hecht, R. M., L.A. Gossett, and W.R. Jeffery. 1981. Dev. Biol. :373-379.Hemmingsen, 5.M., C. Woolford, S.M. van der Vies, K.Tilly, D.T. Dennis and C.P. Georgopoulos,R.W. Hendris and R.J. Ellis. 1988. Nature. .a.:330-334.Henikoff, S. 1984. Gene Z.:351-359.Hershko A. and A. Ciechanover. 1986. Prog. Nuc. Acid Res. 3:1 9-55.Hershko, A. 1988. J. Biol. Chem. a:15237-15240.Hightower, L.E. 1980. J. Cell. Physiol. 1Q2.:407-427.Hiromi, Y., H. Okamoto, W.J. Gehring andY. Hotta. 1986. Cell 44:293-301.Hochstrasser, M., M.J. Ellison, V. Chau, and A. Varshavsky. 1991. Proc. NatI. Acad. Sd. USAL.:4606-461 0.Hockertz, M. K., I. Clark-Lewis, and E.P.M. Candido. 1991 FEBS Lett. 2.i 375-378.Hoffman, E. and V. Corces. 1986. Mol. Cell Biol. 663-673.Hoffman, E.P., S.L. Gerring and V.G. Corces. 1987. Mol. Cell. Biol. Z:973-981.Huang, X.-Y., and D. Hirsh. 1989. Proc. NatI. Acad. Sci. USA :8640-8644.Huitmark, D., R. Klemenz , and W.J. Gehring. 1986. Cell 44: 429-438.127ReferencesIngolia, T.D. and E.A. Craig. 1982. Proc. NatI. Acad. Sd. USA Za:2360-2364.Ireland, R. C. and E.M. Berger. 1982. Proc. Nati. Acad. Sci. USA Zi 855-859.Jentsch, S., J.P. McGrath and A. Varshavsky. 1987. Nature 2.9.: 131-134.Jones, D., D.K. Dixon, R.W. Graham, and E.P.M. Candido. 1989. DNA. :481 -490.Jones, D., R.H. Russnak, R.J. Kay, and E.P.M.Candido. 1986. J. Biol. Chem. Zfil:12006-12015.Kao, H.-T., 0. Capasso, N. Heintz, and J.R. Nevins. 1985. Mol. Cell. Biol. :628-633.Kassenbrock, C. K., P.D. Garcia, P. Walter, and R.B. Kelly. 1988. Nature 90-93.Kay, R. J., R.J. Boissy, R.H. Russnak, and E.P.M. Candido. 1986. Mol. Cell Biol. :3134-3143.Kay, R.J., R.H. Russnak, D.Jones, C. Mathias and E.P.M. Candido. 1987. Nucleic Acids Res.L.:3723-3741.Kingston, R.E., T.J. Schuetz and Z. Larin. 1987. Mol. Cell. Biol. Z:1530-1534.Klemenz, R., D. Huitmark, and W.J. Gehring. 1985. EMBO J. 4:2053-2060.Kramer, J. M., R.P. French, E.-C. Park, and J.J. Johnson. 1990. Mol. Cell Biol. 1.:2081-2089.Krause, M., and D. Hirsh. 1987. Cell :753-761.Kusukawa, N., T. Yura, C. Veguchi, Y. Akiyawa and K. Ito. 1989. EMBO J. :3517-3521.Laird, P.W. 1989. Trends Genet. fi:204-208.Lafrd, P.W., J.M. Kooter, N. Loosbroek, and P. Borst. 1985. Nuc. Acids Res. ja:4253-4266.Landry, J., P. Chretien, H. Lambert, E. Hickey, and L.A. Weber. 1989. J. Cell Biol.j..Q.:7-15.Laszlo, A., and G.C. Li. 1985. Proc. Natl. Acad. Sci. USA :8029-8033.Lee, H., J.A. Simon, and J.T. Lis. 1988. Mol. Cell Biol. .:4727-4735.Leicht, B. G., H. Biessman, K. Palter, and J.J. Bonner. 1986. Proc. Natl. Acad. Sci. USA :90-94.Lenkinsky, R.E., D.M. Chen, J.D. Glickson, and G. Goldstein. 1977. Biochem. Biophys. Ada 44:126-130.Lepock, J.R., K.-H. Cheng, H. Al-Qysi and J. Kruuv. 1983. Can. J. Biochem. Cell Biol. .t:428-437.128ReferencesLeung, D.W., S. A. Spencer, G. Cachianes, R.G. Hammonds, C. Collins, W.J. Henzel, R. Barnard,M.J. Waters and W.l. Wood. 1987. Nature :537-543.Levinger, L. and A. Varshavsky. 1982. Cell Z..:375-385.Li, SC. and Z. Werb. 1982. Proc. Nail. Acad. Sd. USA Z:3218-3222.Lindquist, S. 1981. Nature 2a:311-314.Lindquist, S. and B. DiDomenico. 1985. Coordinate and non-coordinate gene expression duringheat shock. A model for regulation. Academic Press, New York.Lindquist, S. and E.A. Craig. 1988. Annu.Rev. Genet. 22:631-677.Lindquist, S.L. 1980. J. Mol. Biol. 1.Z:151-158.Littlewood, T.D., D.C. Hancock and G.l. Evan. 1987. J. Cell Sc :65-72.Loomis, W.F. and S.A. Wheeler. 1982. Dev. Biol. :412-418.Lund, P.K., B.M. Moats-Staats, J.G. Simmons, E. Hoyt, A.J. D’Ercole, F. Martin, and J.J. VanWyk. 1985. J. Biol. Chem. Z6..:7609-7613.Madura, K., S. Prakash and L. Prakash. 1990. Nuc. Acids Res. la:771-778.Maniak, M. and W. Nellen. 1988. Mol. Cell. Biol. a:153-159.Maniatis, T., E.F. Fritsch and .J.. Sambrook. 1982. Molecular Cloning: A Laboratory Manual.Cold Spring Harbor Laboratory Press.Marota, F.G. and J.M. Sierra. 1988. J. Biol. Chem. Z:15720-15725.Marota, F.G. and J.M. Sierra. 1989. Mol. Cell. Biol. .:2181-2190.Mayrand, S. and T. Pederson. 1983. Mol. Cell. Biol. 161-171.McGarry, T. J. and S. Lindquist. 1985. Cell 42:903-911.McGarry, T.J. and S. Lindquist. 1986. Proc. Nail. Acad. Sci. USA :399-403.Mckenzie, S.L., S. Henikoff and M. Meselson. 1975. Proc. Nail. Acad. Sci. USA ZZ:1117-1121.Mellon, D.A., ed. 1988. Current communications in molecular biology. Antisense RNA and DNA,pp. 71-78. Cold Spring Harbour Laboratory, Cold Spring Harbour, New York.Milkman, R. 1966. Biol. Bull. j1:331-345.Miller, D. 1989. New Scientist. 1:47-50.Mitchell, H.K., G. Moller, N.S. Petersen and L. Lipps-Sarmiento. 1979. Dev. Genet. 1:181-192.129ReferencesMoerman, D. G. and D.L. Baillie. 1979. Genetics J.:95-104.Mondovi, B., R. Stron, G. Rotilio, A.F. Argo, R. Cavaliere and A. Rossi Fanelli. 1969. Eur. J.Cancer :129-136.Moran, L., M.-E. Mirault, A.P.Arrigo, M. Goldschmidt-Clermont, and A.Tissieres. 1978. Phil.Trans. R. Soc. Lond. B. :391-406.Muhich, M.L. and J.C. Boothroyd. 1988. Mol. Cell. Biol. a:3837-3846.Muller, W.U., G.C. Li and L.S. Goldstein. 1985. mt. J. Hyperthermia. 1:97-102.Munro, S. and H.R.B. Pelham. 1986. Cell 4fi:291-300.Munro, S. and H.R.B. Pelham. 1987. Cell :899-907.Neves, A. M., I. Barahona, L. Galego, and C. Rodrigues-Pousada. 1988. Gene : 87-96.Nevins, J.R. 1982. Cell :913-919.Newport, J. and M. Kirshner. 1982. Cell 3..:675-686.Nguyen, V.T., M. Morange and 0. Bensaude. 1989. J. Biol. Chem. 4:1 0487-10492.Nickel, B.E., and J.R. Davie. 1989. Biochemistry :964-968.Nickells, R. W., L.W. Browder and T.l. Wang. 1989. Biochem. Cell BioL ZL687-695.Nickells, R.W. and L.W. Browder. 1988. J. Cell Biol. 1QZ:1 901-1 909.Nieto-Sotelo, J., G. Wiederrecht, A. Okuda and CS. Parker. 1990. Cell 2.:807-817.Nover, L., K.-D. Scharf, and D. Neumann. 1989. Mol. Cell Biol. a:1298-1308.Oppermann, H., W. Levinson and J.M. Bishop 1981. Proc. Natl. Acad. Sci. USA Za:1 067-1071.Otaka, E., K. Higo, and T. Itoh. 1984 . Molec. Geri. Genet. 12:544-546.Ozkaynak, E., D. Finley and A. Varshavsky. 1984. Nature .iZ: 663-666.Qzkaynak, E., D. Finley, M.J. Solomon and A. Varshavsky. 1987. EMBO J. fi: 1429-1439.Padgett, R.A., M.M. Konarska, P.J. Grabowski, S.F. Hardy and P.A. Sharp. 1984. Science22E: 898-903.Parag, H.A., B. Raboy, and R.G. Kulka. 1987. EMBO J. :55-61.Parsell, P.A., Y. Sanchez, J.D. Stitzel and S. Lindquist. 1991. Nature :270-273.Pelham, H. R. B., in J.K. Setlow (ed.) 1987. Genetic Engineering. :27-44. Plenum Press.130ReferencesPeiham, H.R.B. 1982. Cell a.Q.:517-528.Peiham, H.R.B. 1988. Nature 3.32.:776-777.Peiham, H.R.B. and M. Bienz. 1982. EMBO J. 1:1473-1477.Perisic, 0., H. Xiao, and J.T. Lis. 1989. Cell :797-8O6.Perry, G., R. Friedman, G. Shaw and V. Chau. 1987. Proc. Natl. Acad. Sci. USA 4: 3033-3036.Petersen, N.S. and H.K. Mitchell. 1981. Proc. Nati. Acad. Sci. USA Za:1708-1711.Petersen, N.S., G. Moller and H.K. Mitchell. 1979. Genetics :891 -902.Petersen, R.B. and S. Lindquist. 1989. Cell Reg. 1:135-149.Plesofsky-Vig, N. and R. Brambl. 1990. J. Biol. Chem. .:15432-15440.Rabindran, S.K., G. Giorgi, J. Cbs and C. Wu. 1991. Proc. NatI. Acad. Sci. USA :6906-691 0.Reading, D.S., R.L. Hallberg and A.M. Myers. 1989. Nature aaZ: 655-659.Redman, K.L. and M. Rechsteiner. 1989. Nature :438-440.Reymond, C.D. 1987. Nuc. Acids Res. 1:8118.Riddihough, G. and H.R.B. Pelham. 1986. EMBO J. :1 653-1 658.Ritossa, F.M. 1962. Experientia la:571 -573.Rocamora, N. and N. Agell. 1990. Biochem J. 2.Z: 821-829.Rougvie, A.E. and J.T. Lis. 1988. Cell 4:795-804.Ruskin, B., A.R. Krairier, T. Maniatis and M.R. Green. 1984. Cell 3.a:317-331.Russnak, R. H. and E.P.M. Candido. 1985. Mol. Cell Biol. :1 268-1 278.Russnak, R. H., D. Jones, and E.P.M. Candido. 1983. Nucleic Acids Res.il:3187-3205.Sadis, S., E. Hickey and L.A. Weber. 1988. J. Cell Physiol. 1fi: 377-386.Sambrook, J., E. F. Fritsch and T. Maniatis. 1989. Molecular CloninQ : A Laboratory Manual.second edition. Cold Spring Harbor Laboratory Press.Sanchez, E. R., D.C. Toft, M.J. Schlesinger, and W.B. Pratt. 1985. J. Biol. Chem.ZQ: 1 2398-1 2401.Sanger, F., S. Nicklen, and A.R. Coulson. 1977. Proc. Nail. Acad. Sci. U.S.A. Z:5463-5467.Scharf, K.-D., S.Rose, W. Zott, F. Schoff and L. Nover. 1990. EMBO J. :4495-4501.131ReferencesSchlesinger, M.J. 1990. J. Biol. Chem. 2:12111-12114.Schuetz, T.J., G.J. Gallo, L. Sheldon, P.Tempst and R.E. Kingston. 1991. Proc. Nail. Acad. Sci.USA :6911-6915.Scott, M.P. and M.L. Pardue. 1981. Proc. Nail. Acad. Sd. USA Z:3353-3357.Seufert, W. and S. Jentsch. 1990. EMBO J. a:543-550.Shankin, J., M. Jabben, and R.D. Viestra. 1987. Proc. Nail. Acad. Sci. USA :359-363.Shuey, D. J. and C.S. Parker. 1986. J.Biol.Chem. 2fi1:7934-7940.Siegelman, M., M.W.Bond, W. M. Gallatin, T. St. John, H.T. Smith, V.A. Fried and l.L. Weissman.1986. Science Zai.:823-829.Simon, M.C., K. Kitchener, H.-T. Kao, E. Hickey, L. Weber, R. Voellmy, N. Heintz and J.R.Nevins. Mol. Cell. Biol. 1:2884-2890.Skowyra, 0., C. Georgopoulos and M. Zylicz. 1990. Cell 2:939-944.Snutch, T. P. and D.L. Baillie. 1983. Can. J. Biochem. Cell Biol. fil:480-487.Snutch, T.P. and D. L. Baillie. 1984. Mol. Gen. Genet. 1:329-335.Snutch, T.P., M.F.P. Heschl and D.L. Baillie. 1988. Gene 4:241-255.Sorger, P. K. and H.C.M. Nelson. 1989. Cell :807-81 3.Sorger, P.K. 1990. Cell : 793-805.Sorger, P.K. and H.R.B. Pelham. 1987. EMBO J. :3035-3041.Sorger, P.K. and H.R.B. Pelham. 1988. Cell 4.:855-864.Sorger, P.K., M.J. Lewis and H.R.B. Pelham. 1987. Nature aZ.9.:81-84.Southern, E.M. 1975. J. Mol. Biol. :503-517.Southgate, R., A. Ayme and R. Voellmy. 1983. J. Mol. Biol. i:35-37.Spector, A., R. Chiesa, J. Sredy and W. Garner. 1985. Proc. Nati. Acad. Sci.USA Z:471 2-4716.Stinchcomb, D. T., J.E. Shaw, S.H. Carr, and D. Hirsh. 1985. Mol. Cell. Biol. :3483-3496.Stone, D.E. and E.A. Craig. 1990. Mol. Cell. Biol. J&:1622-1632.Storti, R.V., M.P. Scott, A.Rich and M.L. Pardue. 1980. Cell ZZ:825-834.Subjeck, J.R. and T.T. Shyy. 1986. Am. J. Physiol. C1-C17.132ReferencesSubjeck, J.R., T.Shyy, J.Shen and R.J. Johnson. 1983. J. Cell Biol. aZ:1 389-1 395.Suiston, J. E. 1976. Phil. Trans. R. Soc. Lond. B. Z.Z:287-297.Sulston, J. E. and H.R. Horvitz. 1977. Dev. Biol. fi:110-156.Sulston, J. E., E. Schierenberg, J.G. White and J.N. Thomson. 1983. Develop. Biol. I.Qfi: 64-119.Suiston, J.E. and S. Brenner. 1974. Genetics ZZ: 95-1 04.Sung, P., S. Prakash, and L. Prakash. 1988. Genes and Dev. Z:1476-1485.Sung, P., S. Prakash, and L. Prakash. 1990. Proc. Nail. Acad. Sci. USA Z:2695-2699.Sung, P., E. Berleth, C. Pickart, S. Prakash and L Prakash. 1991. EMBO J. J&:21 87-21 93.Sutton, R.E., and J.C. Boothroyd. 1986. Cell 41:527-535.Sutton, R.E., and J.C. Boothroyd. 1988. EMBO J. Z:1431-1437.Tanaka, K., K. Matsumoto, and A. Toh-e. 1988. EMBO J. Z:495-502.Theodorakis, N.G. and R.l. Morimoto. 1987. Mol. Cell. Biol. Z:4357-4368.Thomas, G. H. and S.C.R. Elgin. 1988. EMBO J. Z:21 91-2201.Thomas, G.P., W.J. Welch, M.B. Mathews and J.R. Feramsico. 1982. Cold Spring Harbour Symp.Quant. Biol. 4...:985-996.Thomas, J.D., R.C. Conrad, and T. Blumenthal. 1988. Cell :533-539.Tilly, K. and G. Georgopoulos. 1982. J. Bacteriol. J.4:1082-1088.Tissières, A., H.K. Mitchell and U.M. Tracey. 1974. J. Mol. Biol. 4:389-398.Todd, J.A., T.J.P. Hubbard, A.A. Travers, and D.J. Ellar. 1985. FEBS Lett. i:209-214.Tomasovic, S.P. 1989. Life Chem. Reports. Z:33-63.Toniolo, D., M. Persico, and M. Alcalay. 1988. Proc. NatI. Acad. Sci. USA fi:851-855.Van der Ploeg, L.H.T., S.H. Giannini and C.R. Cantor. 1985. Science Z:1 443-1446.Van Dongen, G., W.L.M. Geilenkreirchen, J. van Rijn and R. van Wijk. 1986. Exp. Cell Res.i.:427-441.Van Doren, K. and D.Hirsh. 1988. Nature :556-559.Vierling, E., R.T. Nagao, A.E. DeRocher and C.M. Harris. 1988. EMBO J. Z:575-581.133ReferencesVijay-Kumar, S., G.E. Bugg, and W.J. Cook. 1987a. J. Mol. Biol. 194:531-544.Vijay-Kumar, S., G.E. Bugg, K.D. Wilkinson, and W.J. Cook. 1985. Proc. NatI. Acad. Sd. USA.a:3582-3585.Vijay-Kumar, S., C.E. Bugg, K.D. Wilkinson, R.D. Viestra, P.M. Hatfield, and W.J. Cook. 1 987b.J. Biol. Chem. Z:6396-6399.Vitek, M. P. and E.M. Berger. 1984. J.Mol. Biolj.Z.:173-189.Voellmy, R., M. Goldschmidt-Clermont, R. Southgate, A. Tissières, R. Levis and W. Gehring.1981. Cell aa:261-270.Welch, W.J. and J.P. Suhan. 1985. J. Cell Biol. Jj:1198-1211.Welch, W.J. and J.P. Suhan. 1986. J. Cell Biol. 19a.:2035-2053.Welch, W.J. and J.R. Feramisco. 1984. J. Biol. Chem. Z5.:4501-4511.Welch, W.J. and J.R. Feramisco. 1985. Mol. Cell. Biol. fi:1229-1237.Welch, W.J. and L.A. Mizzen. 1988. J. Cell Biol. J.Q.:1117-1130.West, M.H.P. and W.M. Bonner. 1980a. Biochemistry 19:3238-3245.West, M.H.P. and W.M. Bonrier. 1980b. Nuc. Acids Res. :4671 -4680.Westwood, J.T., J. Cbs and C.Wu. 1991. Nature :822-827.White, J. G. and H.R.. Horvitz. 1979. Laser microbeam techniques in biological research.Electro-Optical Systems Design AUG.White, J.G., E. Southgate, J.N. Thomson and S. Brenner. 1986. Philos. Trans. R. Soc. Lond. BBiol. Sc 3.14:1 -340.Wiborg, 0., M.S. Pederson, A. Wind, L.E. Berglund, K.A. Marcker, and J. Vuust. 1985. EMB0 J.4...: 755-759.Wickner, S., J. Hoskins and K. McKenney. 1991. Nature 3:165-167.Wiederrecht, G., D. Seto and C.S. Parker. 1988. Cell 4:841 -853.Winning, R.S. and L.W. Browder. 1988. Dev. Biol. j2.:111-12O.Wright-Sandor, L.G., M. Reichlin and S.L. Tobin. 1989. J. Cell Biol. 1Q..:20O7-2016.Wu, B., and R.l. Morimoto. 1985. Proc. NatI. Acad. Sci. USA :6070-6074.Wu, B., C. Hunt and R. Morimoto. 1985. Mol. Cell. Biol. fi:330-341.134ReferencesWu, C., S. Wilson, B. Walker, I. Dawid, T. Paisley, V. Zimarino and H. Ueda. 1987. ScienceZa:1 247-1 253.Xiao, H. and J.T. Lis. 1988. Science :1139-1142.Xiao, H., 0. Perisic and J.T. Lis. 1991. Cell fi4:585-593.Yarden, Y., J.A. Escobedo, W-J. Kuang, T.L Yang-Feng, T.0. Daniel, P.M. Tremble, E.Y. Chen,M.E. Ando, R.N. Harkins, U. Francke, V.A. Fried, A. UlIrich and L.T. Williams. 1986. Nature32a: 226-232.Yost, H.J. and S. Lindquist. 1986. Cell 4:1 85-1 93.Yost, H.J., R.B. Petersen and S.. Lindquist. 1990. Posttranscriptional regulation of hsps in“Stress Proteins in Biology and Medicine”. Cold Spring Harbour Laboratory Press.Zakeri, Z.F., W.J. Welch and D.J. Wolgemuth. 1990. J. Cell Biol. JJj:1785-1792.Ziemiecki, A., M.-G. Catelli, I. Joab and B. Moncharmont. 1986. Bioc. Biop. R. J.:1 298-1307.Zimarino, V. and C. Wu. 1987. Nature 32Z:727-730.Zimmerman, J.L., W. Petri and M. Meselson. 1983. Cell :1161-117O.135AppendixAPPENDIXTable A. Transgenic strains produced in the present study. PC, laboratory strain designation;va, laboratory allele designation; Ex, extrachromosomal; In, integrated; OR, orange cap tank.STRAIN SELECTION IacZ FUSION CNOTYPE LIQUID N2PCi pPD1 0.41 ubq903 vaExi 2:11 8PC2 pPD1O.41 ubqA9O3 vaEx2 3:13PC3 pPD1 0.41 pPC1 6.48-43 vaEx3 3:11 7PC4 pPD1 0.41 ubqt827 vaEx4 1:189PC5 pRF4 pPCZ1 vaEx5 5:1 11PC6 pRF4 pPCZ1 vaEx6 5:1 1 4PC7 pRF4 pPCZ1 vaEx7 5:115PC8 pRF4 pPCZ1 vaEx8 5:130PC9 pRF4- valni 3:18PCi 0 pRF4 pPCZ1 vaExl 0 3:1 8PC11 pRF4 pPCZ1 vaExll 5:180PC12 pRF4 pPCZ1 vaExi2 -PC13 pRF4 pPC16.48-1 vaExl3 6:110PC14 pRF4 pPC16.48-1 vaExl4 6:1 10PC15 pRF4 pPC16.48-1 vaExl5 6:131PC16 pPD1O.41 pPC16.48-1 vaExl6 3:121PC17 pPD1O.41 ubqPvulL903 vaExl7 1:123PC18 pPD1O.41 ubq938 vaExl8 3:121PC19 pPD1O.41 pHS16.25 vaExl9 3:117PC2O pPD1O.41 pHS16.25 vaEx2O 3:13PC21 pPD1O.41 ubq938 vaEx2l 2:64PC22 pPD1 0.41 ubq938 vaEx22 2:64PC23 pPD1O.41 pHS16.25 vaEx23 3:124PC24 pPD1 0.41 ubq938 vaEx24 4:149PC25 pPD1O.41 pHS16.25 vaEx25 4:149PC26 pPD1 0.41 ubqPvuIL903 vaEx26 2:118PC27 pPD1O.41 pHS16.25 vaEx27 3:108PC28 pPD1 0.41 pHS16.25 vaEx28 3:118PC29 pPD1O.41 pHS16.25 vaEx29 3:118PC3O pPD1 0.41 pHS1 6.25 vaEx3O 3:1 08PC31 pRF4 pPC16.48-1 vaEx3l 3:135PC32 pRF4 pPC16.48-1 vaEx32 6:132PC33 pRF4 pPC16.48-1 vaEx33 6:132PC34 pRF4 pPC16.48-1 vaEx34 6:131PC35 pRF4 ubq938NLS vaEx35 3:125PC36 pRF4 ubq938NLS vaEx36 6:143 (OR)PC37 pRF4 pPC16.48-51 vaEx37 4:107PC38 pRF4 pPC16.48-51 vaEx38 6:173 (OR)PC39 pRF4 pPC16.48-51 vaEx39 4:134PC4O pRF4 pPC1 6.48-51 vaEx4O 4:1 34136AppendixTable A (con’t):STRAIN SELECTION IacZ’FUSION (t’)DV,PE LIQUID N2PC41 pRF4 pPC16.1-48XBAI vaEX4l 6:234 (OR)PC42 pRF4 pPC16.41-51 vaEx42 6:173 (OR)PC43 pRF4 pPC16.1-48 vaEx43 2:155PC44 pRF4 pPC16.41-51 vaEx44 2:106PC45 pRF4 pPC16.41-51 vaEx45 3:104PC46 pRF4 pPC16.41-51 vaEx46 2:153PC47 pRF4 pPCZ1 vaEx47 4:X8PC48 pRF4 pPCZ1 vaEx48 2:1 53PC49 pRF4 pPCZ1 vaEx49 2:109PC5O pRF4 pPCZ1 vaEx5O 5:143PC51 pRF4 pPC16.1-48 vaEx5l 5:149PC52 pRF4 pPC16.1-48 vaEx52 1:28PC53 pRF4 pPC16.1-48 vaEx53 1:28PC54 pRF4 pPC16.1-48 vaEx54 1:184PC55 pRF4 pPC16.1-48 vaEx55 1:184PC56 pRF4 pDX16.31 vaEx56 6:108 (OR)PC57 pRF4 pDX1 6.31 vaEx57 6:120 (OR)PC58 pRF4 pDX1 6.31 vaEx58 6:26 (OR)PC59 pRF4 pDX16.31 vaEx59 6:120 (OR)PC6O pRF4 pDX1 6.31 vaEx6O 6:186 (OR)PCG1 pRF4 pDX1 6.31 vaEx6l 6:186 (OR)PC62 pRF4 pPC16.1-48XBAI vaEx62 6:234 (OR)PC63 pRF4 pDX1 6.31 vaEx63 -PC64 pRF4 pPC16.1-48XBAI vaEx64 6:135 (OR)PC65 pRF4 pDX16.31 vaEx65 6:112 (OR)PC66 pRF4 pDX16.31 vaEx66 6:112 (OR)PC67 pRF4 pPC16.1-48XBAI vaEx67 6:135 (OR)PC68 pRF4 pDX1 6.31 vaEx68 6:25 (OR)PC69 pRF4 pPCZ1 valn2 6:228 (OR)PC7O pRF4 pPCZ1 valn3 6:139 (OR)PC71 pRF4 pPCZ1 valn4 6:139 (OR)137

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0086681/manifest

Comment

Related Items